Gaurang Mangla.
I'm a Computer Engineering student at Thapar Institute of Engineering and Technology who builds complete, working systems — agentic AI products with a real frontend and backend, deep learning pipelines trained on data I sourced myself, and enterprise product specs. I don't stop at a notebook — I build the whole thing, spec to shipped product.
Who I am
I build things that ship, not just things that run in a notebook. Whether it's an agentic AI product with a real frontend and backend, a deep learning system trained on data I sourced and cleaned myself, or an enterprise AI platform spec'd out for a Fortune 500 consulting firm — I take ideas from raw data to something a real person can open and use, and I sweat the details that make it feel finished.
Most recently, I worked as a Technology Intern at McKinsey & Company (McKinsey Tech), on the Engagement Lifecycle team, helping define an enterprise Generative AI platform built on the Model Context Protocol (MCP) — translating business requirements into implementation-ready specifications for a platform meant to let business users query governed enterprise data in natural language.
Always open to conversations about AI products, data-driven decision-making, and building things that ship.
What I do
Full-stack product development
Ship complete applications, not just models — React/Vite frontends with FastAPI backends, deployed live end-to-end.
Agentic AI & LLM systems
Multi-node LangGraph pipelines orchestrating search, retrieval, and generation into one coherent product.
Applied ML & deep learning
From CatBoost and Scikit-learn to LSTM/GRU/CNN-LSTM architectures — optimized for the metric that actually matters.
Data engineering
Sourcing and cleaning my own datasets when public ones don't exist — like 8 years of self-scraped AQI data.
Product & business translation
Turning ambiguous business requirements into concrete, implementation-ready specs, wireframes, and workflows.
Database & systems design
Normalized relational schemas, stored procedures, and triggers that enforce integrity at the data layer.
Where I've worked
McKinsey & Company
- Drove product definition for an enterprise Generative AI platform built on Model Context Protocol (MCP), enabling natural-language access to governed enterprise data and eliminating manual dashboard navigation.
- Translated business requirements into implementation-ready specifications across Product, Business Analysis, and UX teams, authoring functional documents, use cases, and user workflows.
- Designed wireframes and prototypes to validate user flows, delivering documentation that enabled engineering to estimate effort and execute the product roadmap.
Tech I work with
Languages
Full-Stack Development
AI / Machine Learning
Data & Tools
Coursework
Beyond the Code
Things I've built
Click any card for the full write-up.
Foundations
Thapar Institute of Engineering and Technology
Patiala, India
Amity International School
Gurugram, India
Continued learning
Machine Learning Specialization
Supervised learning, deep neural networks (TensorFlow), and reinforcement learning.
Fundamentals of Deep Learning
Neural network fundamentals, model training, and deep learning applications.
Let's build something.
Open to conversations about AI products, data-driven decision-making, and building things that ship.