About

I started with maths and physics, and I never really left.

My degree is in AI and Data Science, but the first two years were mathematics and physics. That turned out to be the useful part. The linear algebra is what makes a Kalman filter make sense, and the mechanics is what makes a drone in simulation behave like a body rather than a sprite.

Before any of that I was building model aircraft. I have the aeromodelling certificate to show for it. Flying something you built yourself teaches you quickly that a thing can look correct on the bench and still be wrong in the air.

That is the thread through everything here. I like simulations because they let me build a whole system and then attack it. My best work is usually the moment I found out my own result was not what I thought it was: the clearance check that passed nine times on a broken assembly, the early-warning signal that fired on zero of eight seeds, the deliberately unintelligent baseline that beat everything I had built.

Next is an M.Sc. abroad for 2027, in Computational Science, Scientific Computing, CS or Robotics. I am looking at Germany, the Netherlands, Norway and the USA, and I am learning German for the first of those.

Timeline

  1. 2022

    Started the B.Tech, two years of maths and physics

    The AI and Data Science degree at GGSIPU Delhi opens with mathematics and physics. That is the part I kept using later: linear algebra behind the filters, rigid-body mechanics behind the drone.

  2. 2024

    First research internship: AI for drug discovery

    May to June at GGSIPU. I led the machine learning side for a team of five, building RDKit fingerprints and bioactivity models for colon-cancer compounds. First time a model of mine had to answer to chemistry and not just a loss curve.

  3. 2025

    Prodigal AI, then Product Manager Accelerator

    Summer at Prodigal AI in Delhi on Docker, RAG, OCR and Airflow. From September, backend and ML for StylePilot at Product Manager Accelerator in the USA, remote. The team placed 2nd at the showcase.

  4. 2025 to 2026

    The simulation years: SkyScout, the swarm sim, two papers

    A Mars scout drone with its own CAD, estimator and planner. A drone-swarm threat simulator where a GAN learns to hide attacks. Two preprints, one on grokking predictor signals and one on trust signals for neural PDE surrogates.

  5. May 2026

    B.Tech awarded

    Artificial Intelligence and Data Science, GGSIPU Delhi. Provisional certificate in hand.

  6. 2027

    M.Sc. abroad

    Computational Science, Scientific Computing, CS or Robotics. Germany, Netherlands, Norway, USA. German is at A2 target for the Germany route.

Experience

  1. Backend Developer / ML

    Product Manager Accelerator, USA, remote

    Sep 2025 to Jan 2026
    • Built the StylePilot AI fashion platform: a PostgreSQL schema with 8+ models in SQLAlchemy, and a FastAPI backend.
    • Added pgvector similarity search and a computer vision pipeline using Gemini, SAM and CLIP.
    • Shipped a recommender and a RAG chatbot on top of it.
    • The team placed 2nd at the showcase.
    • PostgreSQL
    • SQLAlchemy
    • FastAPI
    • pgvector
    • Gemini API
    • SAM
    • CLIP
    • RAG
  2. AI/ML Intern

    Prodigal AI, Delhi

    Jul 2025 to Aug 2025
    • Dockerised a startup health-scoring app built in Streamlit.
    • Built RAG and OCR pipelines.
    • Wrote Apache Airflow workflows, in a team of four.
    • Docker
    • Streamlit
    • RAG
    • OCR
    • Apache Airflow
  3. Student Research Intern, AI for Drug Discovery

    GGSIPU, Delhi

    May 2024 to Jun 2024
    • Led the machine learning work in a team of five.
    • Built RDKit molecular fingerprints for candidate compounds.
    • Trained Random Forest and ANN models to predict bioactivity for colon-cancer compounds.
    • RDKit
    • Random Forest
    • ANN
    • scikit-learn