text stringlengths 0 132 |
|---|
--- Page 1 --- |
Aakash R P |
Electronics and Computer Science Engineer |
Trivandrum · Kerala · +91 75610 24795 · aakashrp16@gmail.com · LinkedIn |
Summary |
Results-driven AI Engineer with expertise in Machine Learning, NLP, Computer Vision, and Agentic AI. Skilled in building scalable, |
production-ready systems using Python, LLMs, and multi-agent frameworks. Experienced in integrating APIs, MCP servers, and RAG |
pipelines to deliver end-to-end intelligent solutions across diverse domains. |
Education |
Vellore Institute of Technology (VIT), Chennai |
Sep 2021 – May 2025 |
B.Tech in Electronics and Computer Engineering |
• Current CGPA: 8.09 / 10.0 |
Christ Nagar Central School, Trivandrum |
Jun 2019 – March 2021 |
Senior Secondary (CBSE - 12th Grade) |
• Scored 89.2% overall in board examinations |
Mar Gregorios Memorial Central School, Trivandrum |
Jun 2018 – Mar 2019 |
Secondary School (CBSE - 10th Grade) |
• Scored 95.2% in final board examinations |
Technical Skills |
Languages: Python, R, C, C++, Java, SQL, HTML, MATLAB |
Libraries/Frameworks: TensorFlow, PyTorch, Scikit-learn, Pandas, NumPy, Matplotlib, Seaborn, OpenCV, LSTM, ResNet, FaceNet, |
YOLOv11 |
Tools and Platforms: Git, GitHub, Jupyter Notebook, Vertex AI, Google NLP API, AWS Comprehend, Selenium, SQLite, JIRA API, |
Google Analytics, FastAPI, Flask |
AI/ML and Techniques: Generative AI, Computer Vision, Large Language Models (LLMs), Agentic AI, RAG (Retrieval-Augmented |
Generation),Prompt Engineering, Recommendation Systems, NER, Semantic Matching,Machine Learning, Deep Learning, Natural |
Language Processing (NLP) |
LLM AI Engineering Tools: Claude Code, Claude API, OpenAI API, Gemini (Vertex AI), Groq API, DSPy, MCP (Model Context Protocol), |
A2A Protocol, Skills.md, Guardrails, VectorDB (Qdrant), Tavily, FastAPI |
CS Fundamentals: Data Structures and Algorithms, Object-Oriented Programming (OOP), Operating Systems, Computer Networks, |
DBMS, Compiler Design, Discrete Mathematics, Data Analytics |
Experience |
Alkimi, Bangalore |
Aug 2025 – Present |
AI Engineer |
• Built a multi-agent code migration system using Claude API, automating code analysis and transformation across multiple tech |
stacks. |
• Developed a JIRA monitoring dashboard with live API integration, resolving data bugs and improving performance by 70% via |
per-project caching. |
• Designed A2A protocol documentation and SKILL.md; built a Claude Code session SQLite logger |
• Built and integrated MCP servers connecting internal DSP exchange servers with third-party collaboration and project |
management tools. |
Software Incubator, Trivandrum |
Jun 2025 – July 2025 |
AI Intern |
• Built a three-stage NLP pipeline to automate real estate data extraction and generate AI-driven summaries using Python, SQLite, |
and ChatGPT. |
• Developed scalable Named Entity Recognition (NER) and field mapping solutions using AWS Comprehend, Google NLP API, and |
LLM-based semantic matching to enhance data accuracy. |
UST Global |
Aug 2023 – Nov 2023 |
Machine Learning Intern |
Internship Certificate |
• Led a 2D-to-3D image rendering project using Python, OpenCV, and NumPy, advancing computer vision capabilities for |
visualization tasks. |
• Performed sentiment analysis on large datasets using NLP libraries to extract actionable insights for business intelligence. |
• Designed a machine learning model to predict plot outcomes using supervised learning techniques, improving data-driven |
forecasting. |
--- Page 2 --- |
Projects |
Token Lens | TypeScript, VS Code API, Node.js, Chokidar |
• Built a VS Code extension that monitors Claude Code session logs in real time, parsing JSONL files to surface token usage, cost |
breakdowns, cache efficiency scores, and live event feeds across project and global views. |
• Engineered a prompt analysis engine that detects vague or low-context prompts using pattern matching and generates actionable |
improvement suggestions to reduce exploratory file reads and token spend. |
• Implemented a feature tracker that correlates agent, skill, plugin, and hook usage with isolated token savings, helping developers |
quantify the ROI of Claude Code optimisations. |
Agentic Math Tutor (Agent-RAG) | FastAPI, MCP, DSPy, VectorDB, Guardrails, Groq |
GitHub |
• Built a backend math tutor agent using FastAPI that combines a Qdrant VectorDB knowledge base, Tavily web search, and |
Llama3-based LLM fallback for high-accuracy solutions (80.31% on MathQA). |
• Implemented strict input/output guardrails to block non-math queries and ensure sanitized, math-only responses with cultural |
and ethical filtering. |
• Integrated MCP routing logic and DSPy for human-in-the-loop verification and fallback refinement, creating a safe and |
generalizable Agent-RAG system. |
PRISONSECURE: Smart Prison Surveillance System | YOLOv11, ResNet, FaceNet, Deep Learning |
GitHub |
• Built an AI-driven surveillance system integrating YOLOv11 for weapon detection and ResNet-FaceNet for facial recognition to |
enable real-time monitoring in prison environments. |
• Implemented an alert system that maps detections to prisoner IDs and timestamps, improving incident tracking and response. |
• Achieved high detection accuracy in challenging conditions (low light, occlusion), improving system robustness. |
• Awarded Best Paper at the International Conference on Intelligent Systems and Digital Transformation in collaboration with |
Metropolitan University, Wales. |
ATS-Optimized Resume Analyzer using Gemini Model | Vertex AI, NLP, Generative AI |
• Designed a resume analyzer using Google’s Gemini model to automate and enhance resume screening with keyword matching |
and ranking. |
• Applied Generative AI and custom prompt engineering to extract and score relevant sections of resumes using LLMs. |
• Improved performance using advanced prompt chaining and vector-based similarity search techniques. |
Emotion Detection from Speech | LSTM, MFCCs, Flask, JavaScript |
• Developed a real-time emotion detection model using LSTM networks trained on MFCC features extracted from audio datasets. |
• Built a full-stack interface using Flask (Login, Prediction, Sessions) and an interactive frontend with session history and audio |
visualization. |
• Enabled applications in therapy, human-computer interaction, and customer service by analyzing emotional transitions. |
Web Automation for Record System | Python, Selenium, SQLite |
• Automated data extraction from the Ashland County Recorder portal using Selenium to handle dynamic elements and nested |
End of preview. Expand in Data Studio
No dataset card yet
- Downloads last month
- 10