artifacts/debug_dashboard.md

Top-100 Candidate Discovery -- Executive Triage Dashboard

Re-derived from the real EligibilityEngine / ScoringEngine / features.skills objects via scripts/profile_submission_analytics.py -- no parallel scoring math.

Executive KPI Summary

• Full pool size: 100000
• Top-10 cohort average YOE: 7.74 years
• Active honeypots caught in top 100: 0 / 100 (verified via IntegrityEngine)
• Hard-blocked candidates leaking into top 100: 0 / 100 (should always be 0 -- a nonzero count means the eligible pool fell below 100 and the ranker padded with floored rows)
• Self-learner / dabbler pattern flagged: 43 / 100 (in_career == 0 with trust >= 0.5 on a top-3 contributing group)
• Ghost-profile penalty (behavioral_multiplier < 0.85): 5 / 100

Hard-Gate Fired Rates (full pool, from latest run_report.json)

Gate	Fired	% of Pool
primary_cv_speech_robotics_no_nlp	87,767	87.77%
notice_over_30	54,030	54.03%
pure_research_no_production	26,977	26.98%
outside_india_no_sponsor	21,957	21.96%
title_chaser_sub_18m_hops	18,583	18.58%
consulting_firms_only_career	11,875	11.88%
closed_source_5y_no_validation	7,540	7.54%
outside_experience_band	4,724	4.72%

High-Density Triage Table

Rank	Score	ID	Title @ Company	Top-3 Competency (skill_match)	Logistics & Availability	Gate / Risk Alerts
1	0.504798	CAND_0041610	Anil Subramanian Recommendation Systems Engineer @ Zoho YOE 6.7	llm=0.264 [gpt,langchain,openai] ir=0.235 [embeddings] mlops=0.222 [kubeflow,mlflow] skill_match=0.240168	india_relocatable / sub_30_ideal notice 30d resp 52% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
2	0.503434	CAND_0097176	Advik Mehta ML Engineer @ TCS YOE 5.9	mle=0.403 [pytorch,scikit] nlp=0.222 [nlp,transformer] rank=0.083 [ranking] skill_match=0.236211	preferred_hub / sub_30_ideal notice 30d resp 76% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
3	0.500489	CAND_0074024	Shreya Singh AI Specialist @ Haptik YOE 3.9	mle=0.344 [pytorch,scikit] nlp=0.289 [nlp,transformer] rank=0.083 [ranking] skill_match=0.238921	preferred_hub / sub_30_ideal notice 30d resp 50% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
4	0.498798	CAND_0039754	Mira Banerjee Senior Applied Scientist @ Meta YOE 16.2	rank=0.250 [ranker,ranking,reranking] ir=0.250 [embeddings,faiss,retrieval] retr=0.237 [elasticsearch,retrieval] skill_match=0.245579	india_relocatable / sub_30_ideal notice 30d resp 81% (present)	[✅ CLEAR]
5	0.497311	CAND_0018499	Aarav Trivedi Senior Machine Learning Engineer @ Zomato YOE 7.2	ir=0.484 [embeddings,faiss,retrieval] llm=0.276 [gpt,langchain,llm] rank=0.250 [ranker,ranking,reranking] skill_match=0.336651	preferred_hub / sub_30_ideal notice 15d resp 61% (present)	[✅ CLEAR]
6	0.486131	CAND_0079284	Ishaan Dutta Machine Learning Engineer @ Google YOE 4.9	recsys=0.327 [recommendation] nlp=0.289 [nlp,transformer] mle=0.117 [scikit] skill_match=0.244135	preferred_hub / sub_30_ideal notice 30d resp 79% (present)	[🐥 DABBLER_PATTERN:mle]
7	0.484576	CAND_0037160	Riya Chatterjee Data Scientist @ Haptik YOE 6.0	mle=0.222 [pytorch,scikit] rank=0.083 [ranking] recsys=0.083 [recommendation] skill_match=0.129630	preferred_hub / sub_30_ideal notice 30d resp 74% (present)	[✅ CLEAR]
8	0.484494	CAND_0036437	Arjun Joshi Search Engineer @ Rephrase.ai YOE 4.8	mlops=0.364 [kubeflow,mlflow] rank=0.167 [ranking,relevance] retr=0.091 [elasticsearch] skill_match=0.207020	india_relocatable / sub_30_ideal notice 30d resp 87% (present)	[🐥 DABBLER_PATTERN:retr]
9	0.480946	CAND_0045250	Priya Pandey Applied ML Engineer @ Rephrase.ai YOE 6.6	mlops=0.442 [kubeflow,mlflow] ir=0.121 [embeddings] rank=0.083 [ranking] skill_match=0.215383	preferred_hub / sub_30_ideal notice 15d resp 74% (present)	[🐥 DABBLER_PATTERN:ir]
10	0.477655	CAND_0010770	Ved Mittal Recommendation Systems Engineer @ Aganitha YOE 15.2	rank=0.167 [ranking,relevance] retr=0.083 [elasticsearch] ir=0.083 [faiss] skill_match=0.111111	india_relocatable / sub_30_ideal notice 30d resp 73% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
11	0.477633	CAND_0061257	Advaith Pillai Staff Machine Learning Engineer @ LinkedIn YOE 8.0	rank=0.167 [ranking,relevance] ir=0.058 [retrieval] retr=0.032 [indexing,retrieval] skill_match=0.085623	preferred_hub / sub_30_ideal notice 30d resp 87% (present)	[🐥 DABBLER_PATTERN:ir,retr]
12	0.475440	CAND_0069638	Diya Gupta Computer Vision Engineer @ Swiggy YOE 6.2	nlp=0.222 [nlp,transformer] llm=0.129 [prompt] mle=0.085 [pytorch] skill_match=0.145484	india_relocatable / sub_30_ideal notice 30d resp 64% (present)	[🐥 DABBLER_PATTERN:llm]
13	0.475323	CAND_0068351	Aadhya Iyer Lead AI Engineer @ Sarvam AI YOE 6.4	rank=0.349 [ranking,relevance] nlp=0.066 [nlp] ir=0.059 [retrieval] skill_match=0.157982	preferred_hub / sub_30_ideal notice 0d resp 86% (present)	[🐥 DABBLER_PATTERN:nlp,ir]
14	0.473696	CAND_0008239	Advik Iyer AI Engineer @ Apple YOE 4.0	rank=0.167 [ranking,relevance] retr=0.076 [elasticsearch] llm=0.059 [langchain] skill_match=0.100712	india_relocatable / sub_30_ideal notice 15d resp 73% (present)	[🐥 DABBLER_PATTERN:retr,llm]
15	0.473340	CAND_0027691	Ayaan Goyal NLP Engineer @ Haptik YOE 6.5	mlops=0.324 [kubeflow,mlflow] ir=0.168 [embeddings,faiss] rank=0.167 [ranking,relevance] skill_match=0.219531	preferred_hub / sub_30_ideal notice 15d resp 68% (present)	[✅ CLEAR]
16	0.471628	CAND_0042506	Zara Pandey Search Engineer @ Verloop.io YOE 4.2	nlp=0.299 [nlp,transformer] ir=0.194 [embeddings,faiss,retrieval] llm=0.133 [gpt,openai] skill_match=0.208681	preferred_hub / sub_30_ideal notice 15d resp 48% (present)	[✅ CLEAR]
17	0.470233	CAND_0084283	Riya Naidu Junior ML Engineer @ Sarvam AI YOE 3.6	retr=0.158 [elasticsearch] rank=0.083 [ranking] recsys=0.083 [recommendation] skill_match=0.108290	preferred_hub / sub_30_ideal notice 30d resp 86% (present)	[🐥 DABBLER_PATTERN:retr]
18	0.470126	CAND_0086022	Dhruv Naidu Senior Applied Scientist @ Sarvam AI YOE 5.3	ir=0.437 [embeddings,faiss,retrieval] rank=0.250 [ranker,ranking,reranking] eval=0.222 [mrr,ndcg] skill_match=0.303068	india_relocatable / sub_30_ideal notice 0d resp 55% (present)	[✅ CLEAR]
19	0.469813	CAND_0074123	Karan Sen Data Scientist @ CRED YOE 6.9	nlp=0.303 [nlp,transformer] mle=0.093 [pytorch] rank=0.083 [ranking] skill_match=0.159734	india_relocatable / sub_30_ideal notice 30d resp 38% (present)	[✅ CLEAR]
20	0.468495	CAND_0091534	Dhruv Dutta AI Engineer @ Flipkart YOE 16.6	llm=0.321 [gpt,langchain,openai] rank=0.167 [ranking,relevance] ir=0.167 [embeddings,faiss] skill_match=0.218116	preferred_hub / sub_30_ideal notice 30d resp 84% (present)	[✅ CLEAR]
21	0.468105	CAND_0043860	Pranav Sharma Junior ML Engineer @ Aganitha YOE 6.1	mle=0.203 [pytorch] retr=0.153 [retrieval] ir=0.153 [retrieval] skill_match=0.169975	india_relocatable / sub_30_ideal notice 30d resp 81% (present)	[🐥 DABBLER_PATTERN:retr,ir]
22	0.467548	CAND_0002025	Ira Dalal Senior AI Engineer @ Apple YOE 5.9	nlp=0.299 [nlp,transformer] ir=0.240 [embeddings,faiss] recsys=0.234 [recommendation] skill_match=0.257573	india_non_relocatable / sub_30_ideal notice 30d resp 80% (present)	[✅ CLEAR]
23	0.467277	CAND_0007411	Rahul Bansal Senior Machine Learning Engineer @ Amazon YOE 8.0	llm=0.287 [gpt,prompt] retr=0.243 [retrieval] ir=0.243 [retrieval] skill_match=0.257717	india_relocatable / sub_30_ideal notice 15d resp 12% (present)	[⚠️ GHOST_PENALTY]
24	0.461571	CAND_0081846	Arjun Khanna Lead AI Engineer @ Razorpay YOE 6.7	ir=0.505 [embeddings,faiss,retrieval] rank=0.250 [ranker,ranking,reranking] retr=0.225 [elasticsearch,retrieval] skill_match=0.326946	india_relocatable / sub_30_ideal notice 30d resp 73% (present)	[✅ CLEAR]
25	0.458766	CAND_0057134	Tanvi Joshi Data Scientist @ PhonePe YOE 3.0	mle=0.118 [tensorflow] rank=0.083 [ranking] recsys=0.083 [recommendation] skill_match=0.094966	preferred_hub / sub_30_ideal notice 30d resp 34% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [⚠️ GHOST_PENALTY] [🐥 DABBLER_PATTERN:mle]
26	0.456814	CAND_0068811	Krishna Mittal Applied ML Engineer @ Freshworks YOE 8.0	mlops=0.346 [kubeflow,mlflow] ir=0.220 [embeddings] rank=0.167 [ranking,relevance] skill_match=0.244144	preferred_hub / sub_30_ideal notice 30d resp 42% (present)	[✅ CLEAR]
27	0.447424	CAND_0060054	Aisha Kapoor AI Engineer @ Mad Street Den YOE 6.4	ir=0.370 [embeddings,faiss] recsys=0.233 [recommendation] retr=0.199 [elasticsearch] skill_match=0.267080	india_relocatable / sub_30_ideal notice 15d resp 86% (present)	[✅ CLEAR]
28	0.446938	CAND_0010685	Sunil Mishra NLP Engineer @ Rephrase.ai YOE 6.7	mlops=0.430 [kubeflow,mlflow] ir=0.237 [faiss,retrieval] retr=0.203 [elasticsearch,retrieval] skill_match=0.289681	india_non_relocatable / sub_30_ideal notice 30d resp 83% (present)	[✅ CLEAR]
29	0.446880	CAND_0094759	Aditya Pillai Lead AI Engineer @ Meta YOE 8.6	ir=0.434 [embeddings,faiss,retrieval] llm=0.294 [llm,prompt] rank=0.250 [ranker,ranking,reranking] skill_match=0.325955	preferred_hub / sub_30_ideal notice 30d resp 11% (present)	[⚠️ GHOST_PENALTY]
30	0.446617	CAND_0078492	Aadhya Vora Recommendation Systems Engineer @ Verloop.io YOE 5.1	rank=0.167 [ranking,relevance] ir=0.076 [faiss] mle=0.076 [scikit] skill_match=0.106441	india_relocatable / sub_30_ideal notice 30d resp 70% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:ir,mle]
31	0.446195	CAND_0011687	Shreya Tiwari Senior NLP Engineer @ Niramai YOE 7.8	llm=0.210 [gpt,langchain] ir=0.201 [embeddings,faiss,retrieval] retr=0.083 [retrieval] skill_match=0.164650	india_non_relocatable / sub_30_ideal notice 15d resp 89% (present)	[✅ CLEAR]
32	0.445573	CAND_0022274	Advaith Mukherjee AI Research Engineer @ Yellow.ai YOE 4.5	recsys=0.154 [recommendation] mle=0.111 [scikit] llm=0.057 [prompt] skill_match=0.107247	india_relocatable / sub_30_ideal notice 30d resp 44% (present)	[🐥 DABBLER_PATTERN:recsys,llm]
33	0.445515	CAND_0050454	Saanvi Bansal AI Engineer @ Rephrase.ai YOE 6.8	ir=0.226 [faiss] rank=0.167 [ranking,relevance] retr=0.083 [elasticsearch] skill_match=0.158603	preferred_hub / sub_30_ideal notice 30d resp 77% (present)	[✅ CLEAR]
34	0.444721	CAND_0033179	Yash Subramanian AI Research Engineer @ Wipro YOE 6.9	mle=0.189 [pytorch,scikit] rank=0.083 [ranking] recsys=0.083 [recommendation] skill_match=0.118397	india_relocatable / sub_30_ideal notice 30d resp 84% (present)	[✅ CLEAR]
35	0.444300	CAND_0080051	Arjun Iyer Data Scientist @ Niramai YOE 5.2	rank=0.083 [ranking] recsys=0.083 [recommendation] nlp=0.068 [nlp] skill_match=0.078255	preferred_hub / sub_30_ideal notice 30d resp 54% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:nlp]
36	0.444202	CAND_0027723	Kabir Agarwal ML Engineer @ Wysa YOE 4.2	nlp=0.129 [nlp] rank=0.083 [ranking] recsys=0.083 [recommendation] skill_match=0.098453	india_relocatable / sub_30_ideal notice 30d resp 38% (present)	[🐥 DABBLER_PATTERN:nlp]
37	0.443942	CAND_0087630	Aisha Rao AI Engineer @ Vedantu YOE 7.2	nlp=0.248 [nlp,transformer] ir=0.142 [embeddings] retr=0.104 [elasticsearch] skill_match=0.164595	preferred_hub / sub_30_ideal notice 30d resp 45% (present)	[🐥 DABBLER_PATTERN:retr]
38	0.442857	CAND_0060472	Pari Gupta Computer Vision Engineer @ Paytm YOE 4.4	nlp=0.130 [nlp] mlops=0.127 [mlflow] mle=0.111 [scikit] skill_match=0.122485	india_relocatable / sub_30_ideal notice 30d resp 65% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:nlp,mlops]
39	0.441908	CAND_0052195	Sunil Goyal Computer Vision Engineer @ Sarvam AI YOE 6.3	mle=0.111 [pytorch] rank=0.083 [ranking] recsys=0.083 [recommendation] skill_match=0.092593	india_relocatable / sub_30_ideal notice 30d resp 69% (present)	[✅ CLEAR]
40	0.441840	CAND_0007009	Anika Pillai Recommendation Systems Engineer @ Wysa YOE 7.9	ir=0.356 [embeddings,faiss] llm=0.133 [gpt,openai] retr=0.083 [elasticsearch] skill_match=0.190871	preferred_hub / sub_30_ideal notice 30d resp 62% (present)	[✅ CLEAR]
41	0.437630	CAND_0054394	Ela Iyengar Recommendation Systems Engineer @ PharmEasy YOE 4.1	ir=0.194 [embeddings,faiss] rank=0.167 [ranking,relevance] retr=0.083 [elasticsearch] skill_match=0.147952	preferred_hub / sub_30_ideal notice 30d resp 64% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
42	0.435968	CAND_0043228	Kiara Sen Applied ML Engineer @ Zoho YOE 6.8	mlops=0.367 [kubeflow,mlflow] rank=0.167 [ranking,relevance] nlp=0.142 [nlp] skill_match=0.225269	india_non_relocatable / sub_30_ideal notice 30d resp 41% (present)	[⚠️ GHOST_PENALTY] [🐥 DABBLER_PATTERN:nlp]
43	0.435073	CAND_0061655	Mira Banerjee Machine Learning Engineer @ Krutrim YOE 4.6	nlp=0.282 [nlp,transformer] rank=0.167 [ranking,relevance] recsys=0.083 [recommendation] skill_match=0.177393	india_non_relocatable / sub_30_ideal notice 15d resp 88% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
44	0.434987	CAND_0033445	Ved Bhatia ML Engineer @ Vedantu YOE 6.8	nlp=0.222 [nlp,transformer] mle=0.111 [pytorch] retr=0.105 [retrieval] skill_match=0.146001	india_non_relocatable / sub_30_ideal notice 30d resp 94% (present)	[🐥 DABBLER_PATTERN:retr]
45	0.434329	CAND_0093912	Advik Sethi Senior Data Scientist @ Razorpay YOE 5.3	retr=0.243 [elasticsearch] rank=0.167 [ranking,relevance] ir=0.151 [embeddings,faiss] skill_match=0.186822	india_non_relocatable / sub_30_ideal notice 30d resp 66% (present)	[✅ CLEAR]
46	0.434125	CAND_0049538	Sanjay Bose Applied ML Engineer @ Saarthi.ai YOE 5.8	mlops=0.510 [kubeflow,mlflow] rank=0.167 [ranking,relevance] retr=0.076 [elasticsearch] skill_match=0.250916	india_non_relocatable / sub_30_ideal notice 30d resp 72% (present)	[🐥 DABBLER_PATTERN:retr]
47	0.432840	CAND_0037944	Suresh Singh Senior Data Scientist @ Vedantu YOE 4.9	ir=0.241 [embeddings] llm=0.133 [gpt,openai] rank=0.083 [ranking] skill_match=0.152655	india_non_relocatable / sub_30_ideal notice 30d resp 42% (present)	[✅ CLEAR]
48	0.429882	CAND_0062247	Saanvi Trivedi AI Engineer @ Google YOE 7.3	rank=0.167 [ranking,relevance] retr=0.127 [elasticsearch,retrieval] ir=0.127 [faiss,retrieval] skill_match=0.139999	india_relocatable / sub_30_ideal notice 30d resp 78% (present)	[✅ CLEAR]
49	0.429054	CAND_0051292	Shreya Chatterjee Applied ML Engineer @ Freshworks YOE 5.2	ir=0.363 [embeddings,faiss] retr=0.216 [elasticsearch] rank=0.167 [ranking,relevance] skill_match=0.248589	india_non_relocatable / sub_30_ideal notice 30d resp 52% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
50	0.427753	CAND_0053695	Sanjay Sharma Recommendation Systems Engineer @ Meesho YOE 5.8	ir=0.239 [embeddings,faiss] rank=0.167 [ranking,relevance] retr=0.083 [elasticsearch] skill_match=0.162874	india_non_relocatable / sub_30_ideal notice 15d resp 60% (present)	[✅ CLEAR]
51	0.427459	CAND_0037566	Ritu Nair Machine Learning Engineer @ LinkedIn YOE 6.9	llm=0.318 [gpt,langchain,openai] ir=0.083 [embeddings] retr=0.076 [elasticsearch] skill_match=0.159333	india_relocatable / sub_30_ideal notice 15d resp 50% (present)	[🐥 DABBLER_PATTERN:retr]
52	0.420569	CAND_0025640	Anjali Kapoor AI Research Engineer @ HCL YOE 5.6	nlp=0.222 [nlp,transformer] recsys=0.194 [recommendation] rank=0.083 [ranking] skill_match=0.166471	india_non_relocatable / sub_30_ideal notice 30d resp 87% (present)	[✅ CLEAR]
53	0.420008	CAND_0030031	Anil Joshi AI Engineer @ Microsoft YOE 5.7	nlp=0.369 [nlp,transformer] ir=0.265 [embeddings,retrieval] mlops=0.222 [kubeflow,mlflow] skill_match=0.285263	india_non_relocatable / sub_30_ideal notice 30d resp 94% (present)	[✅ CLEAR]
54	0.419834	CAND_0065878	Suresh Kapoor Senior Data Scientist @ Niramai YOE 7.8	recsys=0.243 [recommendation] rank=0.167 [ranking,relevance] nlp=0.111 [transformer] skill_match=0.173591	india_non_relocatable / sub_30_ideal notice 15d resp 48% (present)	[✅ CLEAR]
55	0.419385	CAND_0079387	Sneha Arora AI Engineer @ Microsoft YOE 6.9	mlops=0.357 [kubeflow,mlflow] recsys=0.336 [recommendation] ir=0.167 [embeddings,faiss] skill_match=0.286724	india_non_relocatable / sub_30_ideal notice 30d resp 81% (present)	[✅ CLEAR]
56	0.418675	CAND_0080534	Manish Iyer ML Engineer @ Genpact AI YOE 3.8	retr=0.148 [elasticsearch] rank=0.083 [ranking] recsys=0.083 [recommendation] skill_match=0.104748	india_non_relocatable / sub_30_ideal notice 30d resp 91% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:retr]
57	0.416002	CAND_0009024	Avni Sharma Search Engineer @ Google YOE 5.2	ir=0.300 [embeddings,faiss] recsys=0.226 [recommendation] nlp=0.111 [transformer] skill_match=0.212158	india_non_relocatable / sub_30_ideal notice 30d resp 46% (present)	[✅ CLEAR]
58	0.414474	CAND_0073675	Anil Goyal Data Analyst @ Ola YOE 6.4	retr=0.129 [retrieval] ir=0.129 [retrieval] nlp=0.058 [nlp] skill_match=0.105463	india_non_relocatable / sub_30_ideal notice 30d resp 73% (present)	[🐥 DABBLER_PATTERN:retr,ir,nlp]
59	0.414383	CAND_0040117	Aisha Sen Recommendation Systems Engineer @ PhonePe YOE 6.5	llm=0.271 [gpt,langchain,openai,prompt] ir=0.220 [embeddings,faiss] nlp=0.111 [transformer] skill_match=0.200813	india_non_relocatable / sub_30_ideal notice 15d resp 66% (present)	[✅ CLEAR]
60	0.412898	CAND_0036863	Vikram Bansal Senior Data Scientist @ upGrad YOE 4.3	llm=0.443 [gpt,openai,prompt] ir=0.222 [embeddings,faiss] rank=0.167 [ranking,relevance] skill_match=0.277052	india_relocatable / buyoutable notice 60d resp 46% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
61	0.412740	CAND_0068964	Avni Malhotra ML Engineer @ Mad Street Den YOE 4.8	nlp=0.352 [nlp,transformer] mle=0.222 [pytorch,scikit] retr=0.032 [elasticsearch] skill_match=0.202032	india_non_relocatable / sub_30_ideal notice 30d resp 70% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:retr]
62	0.408773	CAND_0006418	Rahul Mukherjee Machine Learning Engineer @ Verloop.io YOE 5.7	mlops=0.428 [kubeflow,mlflow] ir=0.181 [embeddings] rank=0.083 [ranking] skill_match=0.230848	preferred_hub / buyoutable notice 60d resp 92% (present)	[🐥 DABBLER_PATTERN:ir]
63	0.408515	CAND_0052682	Ira Mukherjee NLP Engineer @ Aganitha YOE 6.6	ir=0.168 [embeddings,faiss] llm=0.133 [gpt,openai] nlp=0.111 [transformer] skill_match=0.137316	india_non_relocatable / sub_30_ideal notice 30d resp 88% (present)	[✅ CLEAR]
64	0.404030	CAND_0075439	Pooja Mehta Machine Learning Engineer @ Flipkart YOE 4.3	ir=0.183 [embeddings,retrieval] llm=0.133 [gpt,openai] nlp=0.111 [transformer] skill_match=0.142354	india_non_relocatable / sub_30_ideal notice 30d resp 56% (present)	[✅ CLEAR]
65	0.398984	CAND_0015528	Aisha Reddy Applied ML Engineer @ Krutrim YOE 7.4	mlops=0.222 [kubeflow,mlflow] ir=0.195 [embeddings,faiss,retrieval] nlp=0.111 [transformer] skill_match=0.176093	india_non_relocatable / sub_30_ideal notice 30d resp 53% (present)	[⚠️ GHOST_PENALTY]
66	0.397726	CAND_0077337	Aarav Agarwal Staff Machine Learning Engineer @ Paytm YOE 7.0	ir=0.381 [embeddings,retrieval] retr=0.240 [retrieval] recsys=0.228 [recommendation] skill_match=0.282757	india_relocatable / buyoutable notice 60d resp 95% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
67	0.397416	CAND_0036184	Riya Chopra Recommendation Systems Engineer @ CRED YOE 6.0	ir=0.173 [embeddings,faiss] retr=0.083 [elasticsearch] rank=0.083 [relevance] skill_match=0.113327	india_non_relocatable / sub_30_ideal notice 30d resp 90% (present)	[✅ CLEAR]
68	0.395675	CAND_0017590	Sunil Agarwal AI Research Engineer @ Genpact AI YOE 5.1	nlp=0.456 [nlp,transformer] mle=0.256 [pytorch,tensorflow] ir=0.097 [faiss] skill_match=0.269629	india_relocatable / buyoutable notice 45d resp 56% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:ir]
69	0.395627	CAND_0020708	Kiara Patel Search Engineer @ PolicyBazaar YOE 4.2	retr=0.188 [elasticsearch] nlp=0.088 [nlp] rank=0.083 [relevance] skill_match=0.119810	india_non_relocatable / sub_30_ideal notice 30d resp 84% (present)	[🐥 DABBLER_PATTERN:nlp]
70	0.395114	CAND_0058688	Anjali Kapoor AI Engineer @ Vedantu YOE 6.7	mlops=0.222 [kubeflow,mlflow] ir=0.170 [embeddings,retrieval] nlp=0.111 [transformer] skill_match=0.167854	outside_india_no_sponsor / sub_30_ideal notice 15d resp 74% (present)	[⚠️ SOFT:outside_india_no_sponsor]
71	0.394665	CAND_0041669	Aisha Banerjee Recommendation Systems Engineer @ CRED YOE 8.0	llm=0.331 [gpt,openai,prompt] ir=0.204 [embeddings,faiss,retrieval] rank=0.167 [ranking,relevance] skill_match=0.233692	preferred_hub / buyoutable notice 60d resp 77% (present)	[✅ CLEAR]
72	0.394459	CAND_0018549	Mira Verma Recommendation Systems Engineer @ Uber YOE 6.8	retr=0.183 [elasticsearch] rank=0.167 [ranking,relevance] nlp=0.111 [transformer] skill_match=0.153449	india_relocatable / buyoutable notice 60d resp 73% (present)	[🐥 DABBLER_PATTERN:retr]
73	0.394358	CAND_0075574	Karan Ghosh Machine Learning Engineer @ Haptik YOE 5.7	rank=0.167 [ranking,relevance] llm=0.133 [gpt,openai] recsys=0.102 [recommendation] skill_match=0.134118	india_relocatable / buyoutable notice 60d resp 58% (present)	[🐥 DABBLER_PATTERN:recsys]
74	0.393851	CAND_0098846	Shreya Saxena AI Engineer @ upGrad YOE 7.6	mlops=0.346 [kubeflow,mlflow,mlops] rank=0.167 [ranking,relevance] llm=0.133 [gpt,openai] skill_match=0.215298	india_relocatable / buyoutable notice 45d resp 62% (present)	[✅ CLEAR]
75	0.393534	CAND_0070333	Karan Kumar AI Research Engineer @ TCS YOE 4.7	ir=0.174 [faiss] rank=0.083 [ranking] recsys=0.083 [recommendation] skill_match=0.113540	india_relocatable / buyoutable notice 60d resp 48% (present)	[🐥 DABBLER_PATTERN:ir]
76	0.392451	CAND_0082086	Aryan Krishnan Senior Software Engineer (ML) @ Razorpay YOE 6.0	mle=0.256 [pytorch] nlp=0.222 [nlp,transformer] ir=0.133 [faiss] skill_match=0.203929	preferred_hub / buyoutable notice 45d resp 85% (present)	[🐥 DABBLER_PATTERN:ir]
77	0.391200	CAND_0006538	Pooja Malhotra AI Specialist @ Dream11 YOE 5.7	nlp=0.222 [nlp,transformer] mle=0.196 [pytorch,scikit,tensorflow] recsys=0.095 [recommendation] skill_match=0.171148	india_relocatable / buyoutable notice 60d resp 86% (present)	[✅ CLEAR]
78	0.388814	CAND_0048534	Mira Dutta Computer Vision Engineer @ Genpact AI YOE 3.4	nlp=0.222 [nlp,transformer] mle=0.222 [pytorch,scikit] ir=0.029 [embeddings,retrieval] skill_match=0.157704	india_relocatable / buyoutable notice 60d resp 80% (present)	[🐥 DABBLER_PATTERN:ir]
79	0.388281	CAND_0006209	Dhruv Dalal AI Specialist @ Aganitha YOE 3.9	nlp=0.222 [nlp,transformer] mle=0.184 [pytorch] retr=0.126 [retrieval] skill_match=0.177485	preferred_hub / buyoutable notice 45d resp 66% (present)	[🐥 DABBLER_PATTERN:retr]
80	0.386574	CAND_0084681	Ritu Agarwal AI Specialist @ Zomato YOE 3.5	retr=0.167 [retrieval] ir=0.167 [retrieval] rank=0.083 [ranking] skill_match=0.138951	india_relocatable / buyoutable notice 60d resp 67% (present)	[🐥 DABBLER_PATTERN:retr,ir]
81	0.385921	CAND_0067866	Kiara Sethi Senior Software Engineer (ML) @ Tech Mahindra YOE 6.4	nlp=0.222 [nlp,transformer] mle=0.111 [pytorch] rank=0.083 [ranking] skill_match=0.138889	preferred_hub / buyoutable notice 45d resp 79% (present)	[✅ CLEAR]
82	0.385901	CAND_0078262	Amit Joshi ML Engineer @ Yellow.ai YOE 4.9	recsys=0.095 [recommendation] rank=0.083 [ranking] ir=0.049 [embeddings,faiss] skill_match=0.076061	india_relocatable / buyoutable notice 60d resp 39% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:ir]
83	0.385780	CAND_0054703	Sai Shetty Computer Vision Engineer @ Razorpay YOE 6.8	nlp=0.481 [nlp,transformer] mle=0.222 [pytorch,scikit] rank=0.083 [ranking] skill_match=0.262211	india_relocatable / buyoutable notice 60d resp 41% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
84	0.385091	CAND_0046525	Tanvi Mukherjee Senior Machine Learning Engineer @ Genpact AI YOE 6.1	ir=0.546 [embeddings,faiss,retrieval] llm=0.263 [langchain,llm] rank=0.250 [ranker,ranking,reranking] skill_match=0.352977	preferred_hub / buyoutable notice 60d resp 88% (present)	[✅ CLEAR]
85	0.383732	CAND_0060257	Sai Chowdary AI Specialist @ Paytm YOE 5.8	recsys=0.282 [recommendation] mle=0.111 [scikit] rank=0.083 [ranking] skill_match=0.158877	preferred_hub / buyoutable notice 60d resp 57% (present)	[✅ CLEAR]
86	0.382973	CAND_0012837	Myra Chowdary Junior ML Engineer @ Sarvam AI YOE 6.4	mlops=0.151 [mlops] mle=0.111 [pytorch] rank=0.083 [ranking] skill_match=0.115133	india_relocatable / buyoutable notice 60d resp 81% (present)	[🐥 DABBLER_PATTERN:mlops]
87	0.380854	CAND_0034177	Shaurya Sen Senior Software Engineer (ML) @ Mad Street Den YOE 3.6	rank=0.083 [ranking] recsys=0.083 [recommendation] nlp=0.062 [nlp] skill_match=0.076077	india_relocatable / buyoutable notice 60d resp 82% (present)	[🐥 DABBLER_PATTERN:nlp]
88	0.380681	CAND_0011327	Pranav Krishnan AI Research Engineer @ Krutrim YOE 6.3	nlp=0.222 [nlp,transformer] mle=0.111 [pytorch] rank=0.083 [ranking] skill_match=0.138889	preferred_hub / buyoutable notice 60d resp 79% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops]
89	0.380019	CAND_0042029	Dhruv Joshi Senior Data Scientist @ Flipkart YOE 6.5	ir=0.216 [embeddings] nlp=0.171 [nlp] llm=0.133 [gpt,openai] skill_match=0.173614	preferred_hub / buyoutable notice 45d resp 67% (present)	[🐥 DABBLER_PATTERN:nlp]
90	0.379223	CAND_0017648	Arnav Mishra AI Specialist @ Freshworks YOE 4.0	nlp=0.222 [nlp,transformer] mle=0.111 [pytorch] rank=0.083 [ranking] skill_match=0.138889	preferred_hub / over_30_higher_bar notice 120d resp 62% (present)	[⚠️ SOFT:notice_over_30] [⚠️ SOFT:title_chaser_sub_18m_hops]
91	0.379216	CAND_0038099	Sneha Pandey Junior ML Engineer @ Locobuzz YOE 5.4	rank=0.083 [ranking] recsys=0.083 [recommendation] llm=0.074 [prompt] skill_match=0.080357	india_non_relocatable / over_30_higher_bar notice 90d resp 81% (present)	[⚠️ SOFT:notice_over_30] [🐥 DABBLER_PATTERN:llm]
92	0.378712	CAND_0098952	Dev Sethi AI Research Engineer @ CRED YOE 5.5	recsys=0.317 [recommendation] nlp=0.182 [nlp] rank=0.083 [ranking] skill_match=0.194032	preferred_hub / buyoutable notice 45d resp 66% (present)	[🐥 DABBLER_PATTERN:nlp]
93	0.378303	CAND_0081321	Avni Dalal Senior Software Engineer (ML) @ Freshworks YOE 5.3	mle=0.229 [pytorch] nlp=0.222 [nlp,transformer] rank=0.083 [ranking] skill_match=0.178238	india_relocatable / over_30_higher_bar notice 120d resp 47% (present)	[⚠️ SOFT:notice_over_30]
94	0.378048	CAND_0024203	Dhruv Khanna Junior ML Engineer @ Razorpay YOE 4.4	mle=0.263 [pytorch,tensorflow] retr=0.173 [elasticsearch] rank=0.083 [ranking] skill_match=0.172979	india_relocatable / buyoutable notice 45d resp 27% (present)	[⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:retr]
95	0.377870	CAND_0081053	Om Chopra NLP Engineer @ Glance YOE 5.4	mlops=0.222 [kubeflow,mlflow] nlp=0.111 [transformer] recsys=0.083 [recommendation] skill_match=0.138889	india_relocatable / over_30_higher_bar notice 90d resp 83% (present)	[⚠️ SOFT:notice_over_30]
96	0.377837	CAND_0064904	Karan Trivedi AI Engineer @ LinkedIn YOE 4.9	mlops=0.450 [kubeflow,mlflow,mlops] ir=0.321 [embeddings] recsys=0.216 [recommendation] skill_match=0.328850	preferred_hub / over_30_higher_bar notice 90d resp 78% (present)	[⚠️ SOFT:notice_over_30]
97	0.377229	CAND_0007874	Diya Desai Junior ML Engineer @ Sarvam AI YOE 3.7	nlp=0.158 [nlp] mle=0.111 [pytorch] rank=0.083 [ranking] skill_match=0.117580	india_relocatable / buyoutable notice 60d resp 75% (present)	[🐥 DABBLER_PATTERN:nlp]
98	0.376549	CAND_0070589	Aarav Tiwari ML Engineer @ Meesho YOE 3.5	llm=0.143 [langchain] nlp=0.129 [nlp] rank=0.083 [ranking] skill_match=0.118449	india_relocatable / buyoutable notice 60d resp 74% (present)	[🐥 DABBLER_PATTERN:llm,nlp]
99	0.376543	CAND_0044753	Deepak Mukherjee Backend Engineer @ Flipkart YOE 5.6	recsys=0.138 [recommendation] retr=0.091 [retrieval] ir=0.091 [retrieval] skill_match=0.106405	india_relocatable / over_30_higher_bar notice 120d resp 65% (present)	[⚠️ SOFT:notice_over_30] [⚠️ SOFT:title_chaser_sub_18m_hops] [🐥 DABBLER_PATTERN:recsys,retr,ir]
100	0.376483	CAND_0050876	Vivaan Shah Applied ML Engineer @ Freshworks YOE 6.0	ir=0.300 [embeddings,faiss] rank=0.167 [ranking,relevance] nlp=0.111 [transformer] skill_match=0.192481	india_non_relocatable / over_30_higher_bar notice 90d resp 67% (present)	[⚠️ SOFT:notice_over_30]

Model Generation Reasoning Manifest

Rank	ID	Score (6dp)	Reasoning
1	CAND_0041610	0.504798	"Elite alignment -- the work at InMobi is described in the language of someone who built things themselves: a well-grounded hands-on engineering signal, not an architectural or managerial one. Weighed against that, tenure stability is the soft spot here: 50% of roles under 18 months, averaging 20 months apiece, which tempers the otherwise positive read of this profile."
2	CAND_0097176	0.503434	"Exceptional fit -- hybrid retrieval technology visible in the skills: FAISS -- a credible alignment with the JD's hybrid-retrieval requirement. However, tenure stability is the soft spot here: 67% of roles under 18 months, averaging 23 months apiece, which tempers the otherwise positive read of this profile."
3	CAND_0074024	0.500489	"Exceptional fit -- a credible production ML signal from Haptik: not just model training but live inference with the operational responsibility that comes with it. Worth flagging: 50% of recent roles ran under 18 months -- a retention risk the strengths above don't erase on their own."
4	CAND_0039754	0.498798	"Exceptional fit -- hybrid retrieval experience at Apple: the description shows familiarity with the full retrieval stack (dense + sparse), not just one approach -- a marquee fit for what the JD asks."
5	CAND_0018499	0.497311	"Exceptional fit -- Zomato shows up as a retrieval/ranking context -- exactly the core domain this role requires; a best-of-pool case for genuine domain depth rather than keyword proximity."
6	CAND_0079284	0.486131	"Top-tier candidate -- retrieval and ranking domain depth is visible at Swiggy: the work described puts this squarely in the JD's core domain, not just adjacent to it -- a solid signal."
7	CAND_0037160	0.484576	"Standout case -- at Haptik, the engineering context reads as production ML -- deployed systems, live traffic, real operational constraints; a solid answer to the JD's first-order requirement."
8	CAND_0036437	0.484494	"Exceptional fit -- FAISS in the skills section is the JD's hybrid-retrieval signal in concrete form; a dependable match for this specific technical ask."
9	CAND_0045250	0.480946	"Elite alignment -- hybrid retrieval technology visible in the skills: BM25 -- a sound alignment with the JD's hybrid-retrieval requirement."
10	CAND_0010770	0.477655	"Elite alignment -- product-company tenure is one of this JD's core filters; Aganitha in AI/ML clears it -- a high-conviction footing on the most important background requirement. Worth flagging: job-hopping is a real concern -- average tenure runs around 22 months per role, a 50% hop rate."
11	CAND_0061257	0.477633	"Passable fit -- the JD hires for retrieval/ranking expertise built in production; Yellow.ai is where this candidate has done it -- a respectable and specific answer to the role's core ask."
12	CAND_0069638	0.475440	"Passable fit -- a credible production ML signal from Swiggy: not just model training but live inference with the operational responsibility that comes with it."
13	CAND_0068351	0.475323	"Solid fit -- retrieval and ranking domain depth is visible at Sarvam AI: the work described puts this squarely in the JD's core domain, not just adjacent to it -- a competitive signal."
14	CAND_0008239	0.473696	"Solid fit -- the JD is written for someone who has shipped to end users inside a product company; Apple (Consumer Electronics) is exactly that context -- a dependable fit on this screen."
15	CAND_0027691	0.473340	"Workable match -- shipping ML to real users is the hardest signal to fake on a resume, and Haptik's context makes a dependable case that this candidate has actually done it."
16	CAND_0042506	0.471628	"Passable fit -- Verloop.io (Conversational AI) is a product-company context; the career has spent a respectable portion of its time building for users, not billing clients."
17	CAND_0084283	0.470233	"Reasonable case -- Sarvam AI is a solid product-company context; across 1 product role this person has built for end users rather than for clients."
18	CAND_0086022	0.470126	"Passable fit -- product-company tenure is one of this JD's core filters; Sarvam AI in AI/ML clears it -- a credible footing on the most important background requirement."
19	CAND_0074123	0.469813	"Passable fit -- consistently at product companies -- CRED (Fintech) is the kind of org this JD targets, and the tenure there is a premier match for what's being asked."
20	CAND_0091534	0.468495	"Reasonable case -- Glance context is described with vocabulary of someone who codes and ships rather than delegates -- a credible IC signal for a role that values engineering execution over org-chart seniority."
21	CAND_0043860	0.468105	"Solid fit -- hands-on engineering at Aganitha: the description uses builder language -- shipped, built, deployed -- not just 'oversaw' or 'designed'; a credible IC execution signal for a role that needs exactly this."
22	CAND_0002025	0.467548	"Reasonable case -- product-company background at Apple (Consumer Electronics) -- the JD explicitly screens for this context, and a competitive fraction of this career is product-company tenure."
23	CAND_0007411	0.467277	"Passable fit -- shipping ML to real users is the hardest signal to fake on a resume, and Yellow.ai's context makes a robust case that this candidate has actually done it. That said, a 12% recruiter response rate is the friction point here -- most cold outreach goes unanswered, which changes the sourcing approach needed."
24	CAND_0081846	0.461571	"Solid fit -- Razorpay context is described with vocabulary of someone who codes and ships rather than delegates -- a respectable IC signal for a role that values engineering execution over org-chart seniority."
25	CAND_0057134	0.458766	"Passable fit -- domain match on retrieval/ranking: the PhonePe experience puts this in the JD's sweet spot rather than at the margins. The one caveat: retention risk shows up clearly in the tenure data: shortest role at 12 months, average at 18, hop rate at 50% -- all worth surfacing before a final call."
26	CAND_0068811	0.456814	"Good alignment -- the Freshworks description shows evaluation-aware engineering -- ranking metrics are referenced, differentiating this from candidates who tune systems by intuition alone."
27	CAND_0060054	0.447424	"Workable match -- the JD hires for retrieval/ranking expertise built in production; Mad Street Den is where this candidate has done it -- a best-in-class and specific answer to the role's core ask."
28	CAND_0010685	0.446938	"Workable match -- domain match on retrieval/ranking: the Mad Street Den experience puts this in the JD's sweet spot rather than at the margins."
29	CAND_0094759	0.446880	"Good alignment -- production ML track record at Apple: the work described goes beyond experimentation into live systems with real-user stakes -- a competitive signal for this role's primary bar. Even so, 11% response rate to recruiter messages -- a practical sourcing obstacle that warrants a warm intro or a different reach channel rather than cold outreach."
30	CAND_0078492	0.446617	"Reasonable case -- FAISS in the skills section is the JD's hybrid-retrieval signal in concrete form; a robust match for this specific technical ask. The one caveat: job-hopping is a real concern -- average tenure runs around 30 months per role, a 50% hop rate."
31	CAND_0011687	0.446195	"Reasonable case -- product-company background at Niramai (HealthTech AI) -- the JD explicitly screens for this context, and a competitive fraction of this career is product-company tenure."
32	CAND_0022274	0.445573	"Workable match -- evaluation rigour shows up explicitly: Information Retrieval is listed as a skill, and the JD specifically calls for proper offline/online evaluation -- a strong signal this isn't a vibes-based engineer."
33	CAND_0050454	0.445515	"A fair match -- Rephrase.ai (AI/ML) is a product-company context; the career has spent a top-bracket portion of its time building for users, not billing clients."
34	CAND_0033179	0.444721	"A fair match -- the JD's primary bar is production ML; the Wipro context in this profile clears it -- the work is framed in deployment terms, not research terms."
35	CAND_0080051	0.444300	"Workable match -- Niramai shows up as a context where ML went to production, not just to a notebook -- the description puts this in a well-rounded tier for what the role is actually asking for. Weighed against that, the shortest stint on record ran just 14 months, and with a 50% hop rate overall, retention is a fair question to raise before extending an offer."
36	CAND_0027723	0.444202	"A fair match -- Wysa shows up as a context where ML went to production, not just to a notebook -- the description puts this in a sound tier for what the role is actually asking for."
37	CAND_0087630	0.443942	"Good alignment -- Vedantu context is described with vocabulary of someone who codes and ships rather than delegates -- an above-average IC signal for a role that values engineering execution over org-chart seniority."
38	CAND_0060472	0.442857	"A fair match -- product-company background at Paytm (Fintech) -- the JD explicitly screens for this context, and a strong fraction of this career is product-company tenure. That said, the tenure curve here -- 14 months at the shortest, 26 on average -- puts the hop rate at 50% and is the main thing standing between a strong profile and an easy yes."
39	CAND_0052195	0.441908	"Reasonable case -- shipping ML to real users is the hardest signal to fake on a resume, and Sarvam AI's context makes a dependable case that this candidate has actually done it."
40	CAND_0007009	0.441840	"Solid fit -- Wysa is a top-shelf product-company context; across 3 product roles this person has built for end users rather than for clients."
41	CAND_0054394	0.437630	"Good alignment -- the JD asks for hybrid retrieval; the PharmEasy context suggests an above-average hands-on exposure to both dense and sparse methods rather than expertise in one only. The one caveat: job-hopping is a real concern -- average tenure runs around 24 months per role, a 50% hop rate."
42	CAND_0043228	0.435968	"Reasonable case -- the JD asks for hybrid retrieval; the Yellow.ai context suggests a well-rounded hands-on exposure to both dense and sparse methods rather than expertise in one only."
43	CAND_0061655	0.435073	"Passable fit -- evaluation framework usage in evidence at Google: the role description references ranking metrics (NDCG, MRR, or similar), which is the rigour signal the JD is hiring for. Still, job-hopping is a real concern -- average tenure runs around 27 months per role, a 50% hop rate."
44	CAND_0033445	0.434987	"Good alignment -- product-company tenure is one of this JD's core filters; Vedantu in EdTech clears it -- a top-tier footing on the most important background requirement."
45	CAND_0093912	0.434329	"Passable fit -- consistently at product companies -- Razorpay (Fintech) is the kind of org this JD targets, and the tenure there is a high-conviction match for what's being asked."
46	CAND_0049538	0.434125	"Workable match -- Saarthi.ai (Voice AI) is a product-company context; the career has spent a flagship portion of its time building for users, not billing clients."
47	CAND_0037944	0.432840	"Solid fit -- Vedantu (EdTech) is a product-company context; the career has spent a respectable portion of its time building for users, not billing clients."
48	CAND_0062247	0.429882	"Good alignment -- Google is a solid product-company context; across 2 product roles this person has built for end users rather than for clients."
49	CAND_0051292	0.429054	"Passable fit -- Freshworks is a standout product-company context; across 3 product roles this person has built for end users rather than for clients. Weighed against that, job-hopping is a real concern -- average tenure runs around 20 months per role, a 67% hop rate."
50	CAND_0053695	0.427753	"Passable fit -- hands-on engineering at Meesho: the description uses builder language -- shipped, built, deployed -- not just 'oversaw' or 'designed'; a respectable IC execution signal for a role that needs exactly this."
51	CAND_0037566	0.427459	"Workable match -- the work at LinkedIn is described in the language of someone who built things themselves: a credible hands-on engineering signal, not an architectural or managerial one."
52	CAND_0025640	0.420569	"A fair match -- production ML track record at HCL: the work described goes beyond experimentation into live systems with real-user stakes -- a competitive signal for this role's primary bar."
53	CAND_0030031	0.420008	"Passable fit -- core domain evidence at Microsoft -- ranking and retrieval work in a production context; a well-rounded fit for this role's most specific technical requirement."
54	CAND_0065878	0.419834	"Reasonable case -- shipping ML to real users is the hardest signal to fake on a resume, and Uber's context makes a robust case that this candidate has actually done it."
55	CAND_0079387	0.419385	"Reasonable case -- Microsoft (Software) is a product-company context; the career has spent a top-shelf portion of its time building for users, not billing clients."
56	CAND_0080534	0.418675	"Workable match -- the work at Genpact AI is described in the language of someone who built things themselves: a respectable hands-on engineering signal, not an architectural or managerial one. However, 22 months is the average stay here, and at a 50% hop rate this looks more like a pattern than a one-off job change."
57	CAND_0009024	0.416002	"Good alignment -- domain match on retrieval/ranking: the Google experience puts this in the JD's sweet spot rather than at the margins."
58	CAND_0073675	0.414474	"Solid fit -- Ola context is described with vocabulary of someone who codes and ships rather than delegates -- a sound IC signal for a role that values engineering execution over org-chart seniority."
59	CAND_0040117	0.414383	"Solid fit -- PhonePe shows up as a context where ML went to production, not just to a notebook -- the description puts this in a well-rounded tier for what the role is actually asking for."
60	CAND_0036863	0.412898	"A fair match -- Wysa context is described with vocabulary of someone who codes and ships rather than delegates -- a strong IC signal for a role that values engineering execution over org-chart seniority. Worth flagging: the tenure curve here -- 6 months at the shortest, 17 on average -- puts the hop rate at 67% and is the main thing standing between a strong profile and an easy yes."
61	CAND_0068964	0.412740	"Solid fit -- product-company background at Mad Street Den (AI/ML) -- the JD explicitly screens for this context, and a strong fraction of this career is product-company tenure. That said, job-hopping is a real concern -- average tenure runs around 28 months per role, a 50% hop rate."
62	CAND_0006418	0.408773	"Solid fit -- Flipkart shows up as a context where ML went to production, not just to a notebook -- the description puts this in a well-rounded tier for what the role is actually asking for."
63	CAND_0052682	0.408515	"Solid fit -- evaluation framework usage in evidence at Aganitha: the role description references ranking metrics (NDCG, MRR, or similar), which is the rigour signal the JD is hiring for."
64	CAND_0075439	0.404030	"Workable match -- the work at Flipkart is described in the language of someone who built things themselves: a dependable hands-on engineering signal, not an architectural or managerial one."
65	CAND_0015528	0.398984	"Solid fit -- Krutrim context is described with vocabulary of someone who codes and ships rather than delegates -- a solid IC signal for a role that values engineering execution over org-chart seniority."
66	CAND_0077337	0.397726	"Solid fit -- shipping ML to real users is the hardest signal to fake on a resume, and Razorpay's context makes a blue-chip case that this candidate has actually done it. Weighed against that, history shows 50% of roles closing inside 18 months (averaging 21 months) -- a pattern that deserves a straight question in the loop, not a quiet pass."
67	CAND_0036184	0.397416	"Reasonable case -- consistently at product companies -- CRED (Fintech) is the kind of org this JD targets, and the tenure there is a well-rounded match for what's being asked."
68	CAND_0017590	0.395675	"Reasonable case -- Genpact AI shows up as a context where ML went to production, not just to a notebook -- the description puts this in a well-rounded tier for what the role is actually asking for. On the downside, a 50% hop rate is a pattern that calls for a direct retention conversation, rather than an assumption that the next stop will be a long one."
69	CAND_0020708	0.395627	"Reasonable case -- at PolicyBazaar, this person was writing code and shipping systems rather than directing or reviewing -- a dependable execution track record for a role that requires IC depth."
70	CAND_0058688	0.395114	"Workable match -- domain match on retrieval/ranking: the Apple experience puts this in the JD's sweet spot rather than at the margins. That said, based in Germany, outside India, with no sponsorship path on file -- a logistics blocker independent of skill fit."
71	CAND_0041669	0.394665	"Workable match -- product-company background at CRED (Fintech) -- the JD explicitly screens for this context, and a marquee fraction of this career is product-company tenure."
72	CAND_0018549	0.394459	"Solid fit -- a well-grounded production ML signal from Flipkart: not just model training but live inference with the operational responsibility that comes with it."
73	CAND_0075574	0.394358	"Solid fit -- hybrid retrieval technology visible in the skills: BM25 -- a competitive alignment with the JD's hybrid-retrieval requirement."
74	CAND_0098846	0.393851	"Passable fit -- upGrad shows up as a retrieval/ranking context -- exactly the core domain this role requires; a well-grounded case for genuine domain depth rather than keyword proximity."
75	CAND_0070333	0.393534	"Passable fit -- PhonePe (Fintech) is a product-company context; the career has spent an above-average portion of its time building for users, not billing clients."
76	CAND_0082086	0.392451	"Good alignment -- Razorpay context is described with vocabulary of someone who codes and ships rather than delegates -- an above-average IC signal for a role that values engineering execution over org-chart seniority."
77	CAND_0006538	0.391200	"Good alignment -- the JD's primary bar is production ML; the Dream11 context in this profile clears it -- the work is framed in deployment terms, not research terms."
78	CAND_0048534	0.388814	"Good alignment -- product-company background at Genpact AI (AI Services) -- the JD explicitly screens for this context, and a competitive fraction of this career is product-company tenure."
79	CAND_0006209	0.388281	"Reasonable case -- the work at Aganitha is described in the language of someone who built things themselves: a sound hands-on engineering signal, not an architectural or managerial one."
80	CAND_0084681	0.386574	"Passable fit -- core domain evidence at Zomato -- ranking and retrieval work in a production context; a dependable fit for this role's most specific technical requirement."
81	CAND_0067866	0.385921	"A fair match -- Freshworks (SaaS) is a product-company context; the career has spent an above-average portion of its time building for users, not billing clients."
82	CAND_0078262	0.385901	"A fair match -- hybrid retrieval technology visible in the skills: FAISS -- a strong alignment with the JD's hybrid-retrieval requirement. The one caveat: the pattern across this career history -- 50% hop rate, 29-month average stay -- reads as title-chasing risk rather than settled progression."
83	CAND_0054703	0.385780	"Solid fit -- hands-on engineering at Razorpay: the description uses builder language -- shipped, built, deployed -- not just 'oversaw' or 'designed'; a strong IC execution signal for a role that needs exactly this. Worth flagging: job-hopping is a real concern -- average tenure runs around 20 months per role, a 50% hop rate."
84	CAND_0046525	0.385091	"Reasonable case -- Information Retrieval in the skills section is exactly the evaluation-metric literacy the JD flags as a differentiator; a capable credibility signal for a ranking role."
85	CAND_0060257	0.383732	"Passable fit -- the JD's primary bar is production ML; the Paytm context in this profile clears it -- the work is framed in deployment terms, not research terms."
86	CAND_0012837	0.382973	"Good alignment -- production ML track record at Sarvam AI: the work described goes beyond experimentation into live systems with real-user stakes -- a competitive signal for this role's primary bar."
87	CAND_0034177	0.380854	"Workable match -- Mad Street Den (AI/ML) is a product-company context; the career has spent a respectable portion of its time building for users, not billing clients."
88	CAND_0011327	0.380681	"Passable fit -- Krutrim shows up as a context where ML went to production, not just to a notebook -- the description puts this in a sound tier for what the role is actually asking for. Even so, tenure stability is the soft spot here: 67% of roles under 18 months, averaging 25 months apiece, which tempers the otherwise positive read of this profile."
89	CAND_0042029	0.380019	"A fair match -- evaluation framework usage in evidence at Flipkart: the role description references ranking metrics (NDCG, MRR, or similar), which is the rigour signal the JD is hiring for."
90	CAND_0017648	0.379223	"A fair match -- at Freshworks, this person was writing code and shipping systems rather than directing or reviewing -- a well-grounded execution track record for a role that requires IC depth. Worth flagging: 16 months is the average stay here, and at a 67% hop rate this looks more like a pattern than a one-off job change."
91	CAND_0038099	0.379216	"Reasonable case -- BM25 in the skills section is the JD's hybrid-retrieval signal in concrete form; a robust match for this specific technical ask. Worth flagging: notice period runs 90 days, well past the 30-day window the role would prefer -- a real but manageable logistics gap."
92	CAND_0098952	0.378712	"Reasonable case -- CRED context is described with vocabulary of someone who codes and ships rather than delegates -- a robust IC signal for a role that values engineering execution over org-chart seniority."
93	CAND_0081321	0.378303	"Good alignment -- the JD is written for someone who has shipped to end users inside a product company; Freshworks (SaaS) is exactly that context -- an outstanding fit on this screen. Even so, at 120 days, notice runs squarely into 'over 30 higher bar' territory -- not a blocker, but a start-date conversation that needs to happen early rather than after an offer is out."
94	CAND_0024203	0.378048	"Passable fit -- InMobi shows up as a context where ML went to production, not just to a notebook -- the description puts this in a well-rounded tier for what the role is actually asking for. On the downside, 50% of recent roles ran under 18 months -- a retention risk the strengths above don't erase on their own."
95	CAND_0081053	0.377870	"Reasonable case -- Glance (AI/ML) is a product-company context; the career has spent an above-average portion of its time building for users, not billing clients. Still, notice period (90 days, 'over 30 higher bar') is the practical catch here -- everything upstream of the offer stage looks clean."
96	CAND_0064904	0.377837	"A fair match -- core domain evidence at Freshworks -- ranking and retrieval work in a production context; an above-average fit for this role's most specific technical requirement. Set against that, the start-date math is the friction point: 90 days of notice classifies as 'over 30 higher bar', which is worth flagging to whoever is planning the onboarding calendar."
97	CAND_0007874	0.377229	"Reasonable case -- Sarvam AI shows up as a context where ML went to production, not just to a notebook -- the description puts this in a well-rounded tier for what the role is actually asking for."
98	CAND_0070589	0.376549	"Good alignment -- production ML track record at Meesho: the work described goes beyond experimentation into live systems with real-user stakes -- a competitive signal for this role's primary bar."
99	CAND_0044753	0.376543	"Solid fit -- most recently at product companies -- Flipkart (E-commerce) is the kind of org this JD targets, and the tenure there is a sound match for what's being asked. Worth flagging: a 50% hop rate is a pattern that calls for a direct retention conversation, rather than an assumption that the next stop will be a long one."
100	CAND_0050876	0.376483	"Reasonable case -- hybrid retrieval technology visible in the skills: FAISS -- a strong alignment with the JD's hybrid-retrieval requirement. On the downside, notice period (90 days, 'over 30 higher bar') is the practical catch here -- everything upstream of the offer stage looks clean."