Open to opportunities

Havish Jadav

@havishjadav

Message

Software engineer building ML and LLM pipelines—forecasting, RAG chatbots, and API backends for real products.

India

Message

What I'm looking for

I’m looking for a team where I can build production APIs and deploy ML/LLM systems—forecasting pipelines, RAG chatbots, and search-backed features—while shipping with cloud tools like Azure and AWS.

I built a two-stage LightGBM forecasting pipeline for AgriIQ at Dexian, using Agmarknet data to predict commodity arrivals and prices across 150+ markets. I automated the end-to-end ML workflow with Python, PostgreSQL, schedulers, and FastAPI, then deployed forecasting services on Azure Web Apps.

At Mkonnekt, I designed an LLM-driven customer intelligence pipeline with LangChain and LangGraph for 5K+ loyalty users, and I implemented 12+ REST APIs with Java and Spring Boot. I also worked on a Python RAG chatbot backend at Exigotech—using Qdrant and SQLite for semantic search and deploying on Azure VM to streamline navigation across 20+ services.

Experience

Work history, roles, and key accomplishments

Software Intern

Dexian

Apr 2026 - Jul 2026 (3 months)

Developed a two-stage LightGBM forecasting pipeline for AgriIQ to predict agricultural commodity arrivals and prices across 150+ Indian markets. Automated the end-to-end ML workflow using Python, PostgreSQL, and FastAPI, achieving 12.89% WMAPE.

Python PostgreSQL fastAPI LightGBM Azure Machine Learning

Software Developer Intern

Mkonnekt

Jan 2026 - Apr 2026 (3 months)

Designed an LLM-driven customer intelligence pipeline using Python, LangChain, and LangGraph to analyze 5K+ loyalty users. Implemented 12+ REST APIs using Java and Spring Boot for e-commerce integrations.

Python Lang Chain LangGraph Java Spring Boot REST APIs AWS

Software Intern

Exigotech

May 2025 - Jul 2025 (2 months)

Engineered a Python-based website chatbot backend using a RAG pipeline deployed on Azure VM. Delivered vector semantic search with Qdrant and SQLite, reducing navigation time by 30%.

Python RAG Azure Qdrant SQLite Chatbot