Bid Intelligence

Meet Rex.He's read all the bids.Have you?

Rex is the AI assistant built into your tender page. Before you award, he's already compared the field, flagged what's missing, and pulled the attachment evidence — so your team can decide with a clearer map of the bid field.

How Rex AI works

Stage 1 · Quick scan

Six bids. Three scopes.
Who actually covered it?

Rex scans the full field in under a minute — price outliers, missing responses, budget gaps, and your past reviews with each sub. No attachment deep-dive. No winner pick.

  • Average, median and spread across the field, so an outlier is obvious before you open a single PDF.
  • One field scan card per company. Scroll the field, and each card opens up into past performance, how their price stacks up, and the notes to raise before you call them.
  • Hit Dig Deeper on any company and Rex moves that one to Stage 2.

Where Rex lands: a map of the field. Not a recommendation.

Always updating. Those summary tabs at the top, What I'm Seeing, Your Bid Expectations, Next Actions and the rest, don't sit still. Every deep dive you run feeds back into them, and anything Rex rewrites gets marked Updated so your team can see what changed.

Stage 2 · Deep dive

Lowest number looks clean.
Is it?

Dig into one bidder. Rex reads the attachments, checks scope alignment, and pulls your history with them — conditions buried on page four included.

  • Tells you what it actually read: how many attachments, how many past jobs, how many of your own reviews.
  • Splits the read into scope alignment, reliability and commercial signal, so you can go straight to the part you care about.
  • Ends on a final take that lays out the trade-off and then hands the call back to you.

Where Rex lands: evidence. Still your call.

Rex maps the field. You award the job.

AI Checks

Tell Rex what you want him to review.

Every shop has its own scar tissue. The exclusion a sub buried that turned into a change order. The quote that showed up without taxes. The scope gap your trade always gets caught on. AI Checks are standing questions your company sets once, and Rex answers them on every job, in their own tab, one answer per check.

  • Set once, answered every time. Your checks ride along on every job, so nobody has to remember to ask.
  • Field-wide or document-deep. Point a check at the whole bid field, or at one bidder's documents during a deep dive.
  • Answers keep pace. A field-wide check gets revisited after every deep dive. If a document changes the answer, Rex updates it and marks it changed instead of leaving a stale read on the page.
  • On top of the analysis, not instead of it. Rex writes the analysis first, then answers your checks against it. A check can't quietly reshape the findings your team relies on, and the usual limits still hold: no naming a winner, no leaving out evidence it found.

Company administrators own the list, since these shape the analysis the whole team reads. How AI Checks work

Chat with Rex

Follow up on anything.

Rex stays in the tender page. Once Stage 1 is done, your team can ask him about a specific bid, a price gap, a scope concern, or a field flag. He answers from the same analysis — not from scratch.

  • Remembers every Stage 1 and Stage 2 finding for this tender
  • Answers from evidence. Won't guess when data is missing.
  • Stays scoped to the job. Construction talk only.
  • Won't pick a winner. Will tell you everything you need to decide.

Talk to RexAI

Live tender context
Why is Bidder C so much lower than the others?

Rex

Bidder C came in 22% below field median, but they've excluded mobilization and left the equipment requirement response blank. After leveling, the gap closes to about 8%. Worth confirming what's actually in scope before that number does any heavy lifting.
Have we used them before?

Rex

You have one completed job with them on record, rated 3.6 / 5. Timeliness and scope alignment both scored low. No published market reviews available. Not a red flag on its own — but I wouldn't let price carry the decision when the history is that thin.

Performance memory

Price is the start of the conversation.

Rex puts price in context, then checks the rest. Your own review history, market context where it exists, and what the attachments actually say. Award the job on the full picture.

Your review history first

Rex pulls from jobs you've completed and reviewed. Specific scores, specific context. Your institutional memory, not a generic average.

Market context, kept anonymous

Published reviews from other companies round out the picture. Aggregate themes and scores only. Rex never names a job or client that isn't yours to name.

Honest when history is thin

If Rex doesn't have enough to form a view, he says so. Fabricated confidence is worse than a gap. You'll know when the signal is limited.

Rex is already in your tender page.

The analysis starts the moment your team opens a bid. No setup. No separate tool. Hit Run Analysis and let Rex read the field.

Rex is for analysis only - all decisions are yours.