Headline

AI/ML Engineer | Production RAG & Agentic Systems | Python · LangChain · FastAPI | @ NeuralVedaX | open to relocate 🇳🇱

About

I didn't start with AI. I started with curiosity about why things break, why systems fail, and how you'd fix them if you actually had to ship something real. That curiosity took me through four years of CS, two internships, and a final year project that turned into a full multimodal rehab application. Along the way I stopped thinking about ML as a research exercise and started thinking about it as an engineering problem: how do you make retrieval actually accurate? How do you cut LLM costs without cutting quality? How do you build something a real user would trust? Now I'm doing exactly that at NeuralVedaX, working on production RAG pipelines, hybrid search, and LLM orchestration systems. The kind of work where the benchmark isn't a paper metric but whether it holds up under real data. On the side I'm building Totem, an Android app that uses physical QR tokens to make you earn back your screen time. Because sometimes the most interesting problems are the ones nobody's solved cleanly yet. I'm looking for what's next: whether that's a strong engineering role, a research-oriented Masters program, or both. 📬 aditya.ambade@outlook.com

Location

Vadodara

Work experience

AI/ML Developer at NeuralVedax, Feb 2026 - Present · 4 mos, Vadodara, Gujarat, India, Working as an AI/ML Intern at NeurlVedaX Techlabs, contributing to the research and development of intelligent systems and AI-powered applications. • Constructed a 35k-edge recommendation graph from 8,000+ business lines using a pipeline of LLM activity extraction and FAISS embeddings. • Achieved high-precision retrieval across 2,000+ regulatory licenses by designing a hybrid search engine (BM25 + FAISS) with cross-encoder reranking and automated A–F relevance grading. Tech: Python, Docker, FAISS, Azure, Streamlit, Neo4j, PostgreSQL, LangChain, Git.

AI Agent Developer at TechMero, May 2025 - Aug 2025 · 4 mos, Vadodara, Gujarat, India, • Automated an end-to-end publishing pipeline using n8n and OpenAI API, reducing WordPress publishing time by 60%. • Built a production Python backend with JWT authentication and multithreading, increasing system throughput by 40% and handling 50+ concurrent requests via async RESTful APIs. Tech: Python, FastAPI, LangChain, Pydantic, n8n, OpenAI API, Next.js

AI Educator at Cognouso, Jul 2024 - Apr 2025 · 10 mos, Vadodara, Gujarat, India, At Cognouso, I play a pivotal role in designing and implementing AI-driven educational programs/tools that empower students and educators to unlock the full potential of artificial intelligence. I focus on making AI accessible and understandable, ensuring that the next generation of innovators is well-equipped to thrive in a rapidly evolving technological landscape.

Education

Bachelor of Technology - BTech, Computer Engineering, ITM SLS Baroda University

Higher Secondary Certificate, Kendriya Vidyalaya No. 1, Vadodara