Tanya Kansal

About Tanya Kansal

I build human-like intelligence into software blending GenAI with real-world applications. With projects like ODIO, Drive-X, and Astro-Tak, I\'ve moved from backend engineering into the world of LLMs, RAG pipelines, and custom chatbot development. Whether it\'s fine-tuning models like Falcon-7B, building semantic search with FAISS, or integrating Azure Document Intelligence for smart data extraction I specialize in making AI useful, scalable, and human-friendly. I speak both prompt and Python, and my real superpower? Translating complex GenAI ideas into working products. Let’s connect and co-create some AI magic!

Experience

1. Sde i

   MRI Software

   Aug 2024 — Present · Gurugram, IN

   At MRI Software, I worked as a Software Development Engineer I, where I built scalable backend systems and intelligent automation pipelines for enterprise-grade real estate workflows. My stack included Python, Flask, SQL, MongoDB, and Azure Document Intelligence, enabling me to extract structured data from documents, sanitize it, and map it into complex hierarchical SQL formats.In addition to core development, I worked on integrating Generative AI into existing workflows. I designed a custom RAG-based chatbot using LangChain, Pinecone, and OpenAI’s GPT-4 APIs, allowing users to query internal knowledge bases with context-aware responses. I implemented embedding-based semantic search, built a Flask API layer, and developed a user-friendly UI for interaction. I also fine-tuned prompt templates and chain logic to improve accuracy and reduce hallucinations.This role allowed me to blend solid backend engineering with advanced GenAI tools creating solutions that were not only technically robust but also deeply user-centric.

Education

• Sardar Vallabh Bhai Patel University of Agriculture and Technology

  Bachelor of Technology - BTech

  2017 — 2021