Skip to content

01 — Selected Work

Things I've built

01Live

Contract Intelligence

Reads hundred-page construction tenders and extracts the commercial terms and required bid submittals in a single upload.

EPC and government tenders run to hundreds of pages, and pulling out the key commercial terms and the full checklist of bid submittals is slow and error-prone. This full-stack tool runs an uploaded tender through Google Gemini with structured-extraction prompts — one pass returns a 20-field commercial summary, another lists every document the bidder must submit with its page and reason. Results persist with a searchable history, and long documents process in the background so the interface never blocks. Tested on real CPWD and NHAI contracts.

  • React
  • Express
  • MongoDB
  • Google Gemini
  • PDF parsing
Live demo
02In progress

Predictive Maintenance for Water Infrastructure

An ML framework to predict failures in water-network assets before they happen — M.Tech research, IIT Madras × L&T.

In-progress M.Tech research building machine-learning models for predictive maintenance of water infrastructure, tied to L&T's Hubballi–Dharwad 24×7 water supply project. Details and results will land here as the work develops.

  • Python
  • Time-series ML
  • SCADA data
03Live

ML Lab

Eight end-to-end ML models applied to IIT Madras coursework domains and productivity workflows — classical classification through deep-learning sequence models and real-time syncing trackers.

A structured learning series building one complete ML model or system per domain — Construction Economics, DSAI in Finance, Construction Quality, Construction Safety, Smart Power Grid, EMS SCADA, Municipal Infrastructure (Document Intelligence), and Skill & Behavior Tracking. Models are trained/deployed in Python/React, utilizing databases like MongoDB Atlas, Supabase, and Firebase.

  • Python
  • React
  • Google Gemini
  • Supabase
  • Firebase
  • MongoDB
  • TensorFlow
Explore models

02 — About

The combination is the point

I sit at the seam between physical infrastructure and software. My base is electrical engineering, but the work hasn't stayed in one lane: at L&T I've helped plan electrical systems and commission the SCADA controls behind a major water-supply scheme, and at IIT Madras I'm researching machine-learning models that predict infrastructure failures before they happen.

Large infrastructure projects generate enormous amounts of data that mostly goes unused. What interests me is turning it into decisions — which asset to service first, which clause in a 300-page tender actually matters, which failure is coming next. My Contract Intelligence and predictive-maintenance work both come from that same instinct.

I'm doing my M.Tech at IIT Madras while still on L&T's rolls, which keeps me close to real projects as I build. I'm drawn to roles and collaborations where engineering domain knowledge and ML actually meet — not one or the other.

  • 2025 — 2027

    M.Tech, Construction Technology & Management

    IIT Madras

    Predictive-maintenance ML research, in collaboration with L&T (Hubballi–Dharwad 24×7 water supply).

  • 2023 — Present

    Graduate Engineering Trainee

    Larsen & Toubro

    Electrical planning and SCADA commissioning on the SMCB water-supply scheme.

  • B.Tech

    Electrical Engineering

    Jabalpur Engineering College

    Foundation in power systems and electrical engineering.

03 — Skills

What I work with

Machine Learning & Data Science

  • Python
  • scikit-learn
  • TensorFlow / Keras
  • LSTM networks
  • Time-series modeling
  • Pandas
  • NumPy
  • Matplotlib
  • Jupyter

Infrastructure & Domain

  • SCADA & industrial control (DNP3, IEC 60870)
  • Power systems (EMS, State Estimation)
  • Electrical substations
  • Smart grids
  • Water infrastructure
  • Predictive maintenance
  • Construction planning & control
  • Contract management

Tools & Platforms

  • Git & GitHub
  • Revit & Dynamo (BIM)
  • Next.js / React
  • MongoDB
  • Google Gemini API
  • Vercel

04 — Contact

Let's build something

I'm enthusiastic for the conversations that lead to research collaborations, where infrastructure engineering and ML meet. The fastest way to reach me is email — or use the form.

ce25m104@smail.iitm.ac.inLinkedInGitHub