Ashwin mendhe

Developed a basic RAG pipeline using LangChain, Groq, ChromaDB, FAISS CPU, and LLaMA 3.1 8B Instant for efficient information retrieval and response generation. Developed and deployed RESTful API Design using FastAPI and Django REST, supporting a system that manages over 100 devices and reducing device provisioning time by 80%. Developed scalable API integration and backend development reducing data processing time by 70% for over 100 users. Enhanced SDLC efficiency, data security, and scalability through JWT token-based authentication and microservices architecture. Implemented integration tests, improved code quality, and reduced production issues by 80%. Transformed unstructured MongoDB data to structured MySQL using Pandas and NumPy. Debugged complex software and device issues on Linux/macOS. Designed and validated OAS specifications, enhancing API documentation clarity. Integrated Gemini 2.0 Flash AI agent for debugging processes. Developed live video streaming systems using FFmpeg, RTMP, and AWS S3. Automated monitoring of 100+ servers and devices using Python plugins in Site24 7. Developed cross-platform SDKs, PyPI packages, CLI tools, and CI/CD pipelines. Built Teams notification system using pymsteams and MongoDB. Managed Python scripts to streamline software development processes.

Developed and deployed CNN and ANN models using scikit-learn for pose estimation and classification, achieving 80% accuracy and reducing processing time by 68% for real-time analysis across 1000+ images. Deployed inference APIs on cloud functions (AWS) utilizing FastAPI/Django and WebSocket, achieving 100% uptime and supporting over 20 real-time data transactions per second.

Developed and deployed CNN and ANN models using scikit-learn for pose estimation and classification, achieving 80% accuracy and reducing processing time by 68% for real-time analysis across 1000+ images. Deployed inference APIs on cloud functions (AWS) utilizing FastAPI/Django and WebSocket, achieving 100% uptime and supporting over 20 real-time data transactions per second.

Developed backend modules for cryptocurrency analytics using Django and PostgreSQL, processing over 1k transactions daily and improving data retrieval speed by 100%. Developed and deployed ETL data pipelines, data modeling utilizing pandas and numpy, and integrating 10+ real-time exchange APIs (e.g., Binance, WazirX), processing over 30-40 transactions daily and reducing data latency by 100%.

Developed 3 Flask-based dashboards, improving data visualization for healthcare analytics, leading to a 100% reduction in report generation time for over 50 users. Automated analysis of 1K+ TCP/UDP/RTP packets using Python scripts, reducing manual analysis time by 90%. Researched and analyzed multimedia pipelines, including GStreamer and FFmpeg, to optimize video transmission efficiency.