VendorLens / ABOUTME.md
Vishwajeet07's picture
initial
b8f6c99

About Me

Vishwajeet Kumar

Software engineer with ~2 years of experience across full-stack, backend systems, and AI/ML integration. I've worked at a logistics startup, an edtech fintech, and done freelance work β€” mostly building things from scratch or scaling things that were already in production. I like working across the whole stack but I'm most comfortable in backend and systems work.

Contact: vishwajeet.7t@gmail.com | +91-9661241777 | Bengaluru, India Links: LinkedIn | GitHub | Portfolio


Education

B.Tech β€” Electrical and Electronics Engineering National Institute of Technology, Sikkim Dec 2021 – May 2025

Relevant coursework: Data Structures & Algorithms, Operating Systems, Machine Learning, Python for Data Science, Microprocessors


Work Experience

Software Development Engineer 1 Logizee Solutions LLP (Dasho) β€” Pune | Jul 2025 – Nov 2025

  • Engineered a scalable logistics platform using TypeScript/Kotlin microservices and Flutter, supporting 3,000+ monthly deliveries on auto-scaling infrastructure
  • Improved ML-based demand forecasting accuracy by 18%, reducing dispatch delays and enhancing real-time operational insights
  • Built a production-grade inverted-index search engine and integrated secure payment rails (Zoho, Razorpay), enabling low-latency transactions
  • Automated CI/CD pipelines with canary deployments and instant rollback, increasing deployment stability and minimising downtime

Software Development Engineer Eduvanz Financing Pvt. Ltd (Wizr) β€” Mumbai | Feb 2025 – Jun 2025

  • Integrated Monster and OpenAI APIs via a TypeScript/Node.js backend with Redis caching and vector search, serving optimised APIs to Next.js and React UIs for 10,000+ users
  • Delivered scalable backend workflow services using FastAPI and Node.js, cutting manual processing by 30% through automated orchestration
  • Accelerated Next.js UI performance through rendering optimisations, code splitting, and lazy loading, achieving 22% faster page loads

Freelance Software Developer Groozo β€” Sikkim | Sep 2024 – Jan 2025

  • Built a real-time, workflow-driven delivery system using Node.js, Flutter, and Firebase, processing 12,000+ orders with clean architecture principles
  • Applied AI-assisted routing with Mapbox optimisations, lowering delivery ETAs by ~25%
  • Improved backend throughput using Redis caching, load-balanced microservices, and monitoring, reducing API latency by 30% under peak load

Junior Embedded and IoT Engineer 9Pointers Tech Pvt. Ltd β€” Jaipur | Jun 2024 – Aug 2024

  • Built a real-time computer-vision access control system with Python, OpenCV, and TensorFlow, reaching 98.5% accuracy and cutting access latency by 40% via device-side optimisation
  • Provisioned IoT control and monitoring services using ESP32, FastAPI, and MQTT, achieving 99.8% command reliability with sub-300ms latency across distributed devices

Skills

Languages: Java, Python, TypeScript, Kotlin, C++, C#

Frontend & UI: React.js, Next.js, React Native (Expo), Flutter, Tailwind CSS, Figma

Backend & APIs: Node.js, Spring Boot, FastAPI, REST APIs, gRPC, JWT Auth

Databases: PostgreSQL, MySQL, MongoDB, Redis, Firebase, Vector Databases

DevOps & Cloud: Git, Docker, Kubernetes, Terraform, Prometheus, Grafana

AI / ML: LLM Integration, RAG Pipelines, Embeddings, Prompt Engineering, LangChain, OpenAI APIs, TensorFlow


Other Projects

X-Ray System β€” Pipeline Decision Debugging Platform (Jan 2026) Multi-tenant async observability system for non-deterministic pipelines (RAG, search, recommendations). Captures 100% step-level decision context across 5,000+ candidates per step with zero production latency impact. FastAPI + PostgreSQL (JSONB) backend with a Python SDK for millisecond-level trace queries.

Tina AI β€” AI Interview Agent Platform (May–Jun 2025) Adaptive screening engine using Next.js, FastAPI, WebRTC, and LLM orchestration. Responsive voice interactions with scenario-driven questioning at <200ms latency. Context-aware evaluation pipeline with 97% scoring consistency and timestamped transcript search.


Achievements

  • Scaled a real-time delivery platform to 12,000+ production orders through performance optimisation, monitoring, and cross-team collaboration
  • Led development of NIT Sikkim's React-based conference website serving 6,000+ users
  • Contributed to FastAPI and Material UI open-source; secured 2nd place at IIT-BHU TechNex