Himanshu Sanwal — AI Researcher

I ship entire SaaS products before lunch. That's not a flex — it's just what happens when you master agentic development. I think in systems. I see a problem, decompose it into services, design the data model, define the API contracts, and orchestrate AI agents to implement the entire stack while I architect and review. What I've built speaks for itself — production ML pipelines with Cython and Fortran-compiled backends for high-performance statistical computing. Real-time web scraping infrastructure that ingests, normalises, and serves data from multiple sources on automated schedules. RAG systems with custom chunking strategies over structured domain data. Conversational AI layers that parse natural language queries, route to the right data tables, and return contextualised answers. Multi-agent architectures where specialised agents research, reason, and act with shared memory and social signalling. Full-stack platforms with Docker-containerised Python backends, TypeScript frontends, PostgreSQL with pgvector for semantic search, Supabase for auth and real-time subscriptions, and CI/CD pipelines deploying to Vercel and cloud infrastructure. My foundation is data science and machine learning — prediction models, feature engineering, model evaluation, data pipelines from raw scrapes to clean APIs. That foundation means I don't just call LLM APIs. I understand tokenisation, attention mechanisms, embedding spaces, and inference trade-offs underneath. But what sets me apart is velocity. Give me a problem on Monday, I'll hand you a deployed product by Tuesday. Auth, database, API layer, frontend, payments, monitoring — live on a real URL. I've done it enough times that my development workflow is a system itself. Tech stack: Python (NumPy, SciPy, scikit-learn, pandas, FastAPI, Cython), TypeScript, React, Next.js, React Native, Node.js, Supabase, PostgreSQL, pgvector, Docker, Vercel, Claude Code, Cursor. AI/ML depth: RAG pipelines, agentic architectures, multi-agent orchestration, MCP protocol, tool/function calling, prompt engineering, vector databases, LLM evaluation frameworks, fine-tuning trade-offs, model serving, cost optimisation. Currently open to AI Engineer roles — remote, full-time, or contract. Looking for a team that measures engineers by what they ship, not how many meetings they attend. Hiring? DMs are open. Building with AI? Follow along — I'm sharing everything I learn, daily.

Stackforce AI infers this person is a SaaS-focused AI Engineer with expertise in Generative AI and Machine Learning.

Experience: 5 yrs 2 mos