About

Building the data + AI layer for
Indian infrastructure

InfraLens makes Indian Standard codes, design rules, BOQ builders and engineering AI free, searchable, and queryable — for the 4M+ civil engineers, contractors and PMCs delivering India's ₹143 lakh crore infrastructure pipeline.

The thesis

Civil engineering has the deepest data, and the least structure

India is delivering its biggest infrastructure build in history — ₹143 lakh crore through 2025, 10,000+ km of highways every year, 100 smart cities, and 416 million more urban residents by 2050. The workflows running all of it haven't fundamentally changed in two decades.

Civil engineering carries the deepest data of any engineering discipline — IS / IRC / NBC codes, CPHEEO manuals, DPR archives, BOQ tenders, QA/QC records, hydrology, geotechnical reports, mix designs. Almost none of it is structured. Almost none of it is queryable. Senior engineers still cross-check one clause across eight scattered sources, while two decades of tacit knowledge walks out the door with every retirement.

InfraLens is building the structured-data + AI layer that fixes this. The free reference layer you see today — codes, handbook, calculators, templates — is the foundation. The AI-driven decision-support stack sits on top, and is what we're shipping toward.

The base layer

We didn't write the codes — we made them queryable

India's engineering standards — IS · IRC · NBC · MoRTH · CPHEEO · state SORs — already exist. They're published by BIS, the Indian Roads Congress, state PWDs and central ministries. The problem isn't that they're missing; the problem is they live as scattered PDFs across two dozen portals. We structured them into a single searchable spine, then built engineering tools on top.

Structured data spine
Tools built on top
What we're building

Engineering AI for India's infrastructure stack

Three active R&D initiatives sit on top of the base layer — each a meaningful slice of the engineering decision loop we're shipping toward.

🔍
Active build
SOR Intelligence — search by intent
  • Indian SOR documents are 500-page PDFs, one per state PWD — thousands of line items locked in long-form specs.
  • We structure every line into queryable fields (material · grade · dimensions · unit · rate).
  • Contractors search by what they need, not by hunting through a hundred-page index.
  • Spec-aware semantic match across 291 state and central SORs.
📐
Active build
Quantity take-off from drawings
  • Every tender starts with a bidder hand-reading drawings and BOQ PDFs — a 2-3 day step before pricing even begins.
  • We extract structured quantities directly from drawing PDFs — walls, columns, slabs, reinforcement.
  • Bidder supplies scope & grade inputs; quantities come back automatically, linked to clauses.
  • Cuts the take-off step from days to minutes — and removes the most error-prone bid stage.
🗺️
Active build
TP Subdivision Engine
  • Every Indian city has its own Town Planning Scheme — plot sizes, road widths, setbacks, FSI, open-space ratios — all different.
  • Drop a raw land-parcel boundary in; the engine generates valid subdivision layouts in seconds.
  • Every layout is compliance-checked against the city's active TP rule set.
  • Replaces a week of architect iteration with a queryable, optimisation-driven layout space.

Working on the same problems?

Research partnerships, data collaborations and early-access invites — write to us with the brief, we read every note within 48 hours.

Get in touch
Founder

Who built this

SA
Sourabh Agarwal
Founder · Infrastructure Analyst

I work at the intersection of infrastructure engineering, capital markets, and data-driven decision systems.

With a background in civil engineering and project execution, I have evolved into a deep-dive infrastructure analyst — decoding EPC companies, InvITs, urban development models, and large-scale public infrastructure programs across India.

Through Infra Lens, I simplify complex engineering and policy frameworks into structured, research-backed insights.

Work through InfraLens
  • Highway & EPC sector analysis
  • InvIT & infrastructure asset valuation
  • Smart city & riverfront development models
  • Water grid & large-scale utility planning
  • DPR structuring & feasibility documentation
Also at StockLens — systematic investment frameworks
  • Earnings transcript & rating-report analysis on listed Indian companies
  • Multi-factor simulation models (technical, fundamental, sentiment)
  • Backtesting of 20+ quant strategies
  • Python automation + HFT simulations
  • Research-grade structured financial databases
Currently exploring
  • Textile manufacturing (Surat-based rapier & water-jet loom setup)
  • Woven geotextile plant development (bank-ready DPR structuring)
  • AI-assisted civil engineering handbook applications
  • Quant-driven capital allocation systems
Connect on LinkedIn
Data & methodology

How the data is built

  • Bureau of Indian Standards (BIS) published code listings as the primary spine.
  • Archive.org and law.resource.org for free PDF availability — 86 % of codes have direct links, the rest fall back to Google search.
  • Automated enrichment (LLM-assisted) extracts key tables, clauses, values, and FAQs from each code.
  • Manual verification of the highest-traffic codes (IS 456, IS 800, IS 1893, IS 875, IS 10262, IS 13920).
  • Quarterly health-check audits of external PDF links — broken URLs are detected and either re-sourced or removed.

Always refer to the original IS code document for design and compliance decisions. InfraLens is not affiliated with the Bureau of Indian Standards.

Contact

Get in touch

Also by us

Other projects

StockLens.co.in →

Systematic Indian-equities research. CEO personality mapping of the Top 1,000 corporate leaders, IAS-officer screener, and corporate-governance insights.

InfraLens is an independent research initiative. The information provided is for educational and informational purposes only. IS code data and engineering-calculator output should not be used as a substitute for reading the original BIS documents or a professional structural-engineering review. We are not affiliated with the Bureau of Indian Standards.