Built for Indian public tenders

Every relevant tender. Read, checked and ready before the deadline.

TenderLens finds tenders for your keywords every night, reads all their documents and checks whether your company qualifies, showing the page behind every condition.

Sign in to TenderLensSee how it works
Nightly
keyword search
Page-level
citations
3 verdicts
with reasons

Sample · State IT department

AI-based face recognition attendance for 3,900 colleges

9 days left
Needs reviewEMD ₹2,50,000RFP · 127 pages
  • Average turnover ≥ ₹5 CrRule engine · FY 2023-26p.14
  • ISO 27001 certifiedValid until May 2028p.17
  • ?3 similar e-governance worksCheck by hand: AI unsurep.20
  • Office in the stateCorrigendum 2 relaxed thisp.23

Every condition links to the page it came from.

Reads: PDF, Scanned PDF · OCR, Hindi + English, Word, Excel, ZIP archives, GeM bid documents, Buyer terms (ATC), Corrigenda, BoQ sheets.

How it works

From a keyword to a bid decision, every morning.

The routine work of reading tenders runs overnight, so your team starts with the ones that matter.

  1. Step 1

    Discover

    Every night it searches for your keywords and their pairs, and keeps the tenders that score high.

  2. Step 2

    Read

    It opens the notice, the RFP, the annexures and the corrigenda, scanned pages included.

  3. Step 3

    Extract

    It pulls out each eligibility condition: turnover, experience, certificates, with the page it came from.

  4. Step 4

    Check

    It compares each condition with your company profile and says Eligible, Needs review or Not eligible, with reasons.

  5. Step 5

    Track

    It watches for corrigenda and new closing dates, and tells you in the morning summary.

Why teams trust it

Careful by design, because a missed tender is a missed opportunity.

Evidence, not guesses

Every condition and every verdict quotes the tender page it rests on, checked against the document text.

Numbers checked by rules

Turnover, net worth, years of profit and dates are checked by a tested rule engine, not by an AI's impression.

When unsure, it asks you

If a page couldn't be read or the AI isn't sure, the tender goes to Needs review. A good tender is never quietly dropped.

Your company data stays private

Tender text and company facts go to separate AI services. Your profile never reaches the public model.

Careful with your paid account

It browses like one careful person, stays within a daily budget, and never buys anything on its own.

On your own server

One self-contained installation on your Linux server, with nightly backups and a tested restore.

Clear answers

Three verdicts. Every one explained.

Each tender gets a verdict and a checklist of its conditions. Open any line to see the tender’s own words, the company fact it was compared with, and the reasoning. If you disagree, override it with a note; re-runs keep your decision.

Eligible

Every mandatory condition is met by your profile, through an exemption, or by a declaration you can make.

Needs review

Something needs a person: a page that couldn't be read, missing data, or an AI judgement worth a second look.

Not eligible

A clear rule isn't met, like a turnover threshold or a required certificate, with the page that says so.

Ready when the next tender drops.

Accounts are created by your administrator. Sign in to see today’s tenders.

Sign in