Breaking
Revere joins Y Combinator's F25 batch out of San Francisco AI agents model competitor pricing for government contractors Core metric: modeled pWin at every price point Founders bet weeks of research can shrink to hours Legal entity: Tig Technologies Inc. Tagline: "Your unfair advantage" Revere joins Y Combinator's F25 batch out of San Francisco AI agents model competitor pricing for government contractors Core metric: modeled pWin at every price point Founders bet weeks of research can shrink to hours Legal entity: Tig Technologies Inc. Tagline: "Your unfair advantage"
Company · GovCon × AI

Revere Thinks Federal Contracts Are Won and Lost on One Number

The Y Combinator-backed startup builds AI agents that price a government bid the way a rival would. Its whole product distills to a probability: how likely are you to win, and at what cost to your margin.

Ask anyone who has bid on a federal contract what keeps them up at night, and the answer usually is not the technical write-up or the compliance matrix. It is the price. Set it too high and a hungrier competitor takes the award. Set it too low and you win work that quietly loses money for years. The number sits at the center of every proposal, and for decades teams have arrived at it the same way: weeks of spreadsheets, a few phone calls, and a good deal of gut feel.

Revere, a startup in Y Combinator's Fall 2025 batch, wants to replace the gut feel with a model. The company builds AI agents that study a contract opportunity the way a rival bidder would - pulling labor rates, past awards, and hiring signals - and then estimates the odds of winning at each price you might submit. Internally the company runs under the name Tig Technologies Inc. To the market it goes by Revere, and its pitch fits on a bumper sticker: your unfair advantage.

The ProblemPricing in the Dark

Government contracting is an enormous market, and a strange one. Buyers are agencies, timelines run long, and the rules reward whoever can offer the most credible work at the lowest defensible cost. But the information a bidder needs to price well is scattered. Comparable awards live in one database. Labor rates live in job postings and public schedules. A competitor's likely cost structure has to be inferred from fragments. Assembling all of it into a coherent view of the field can take a proposal team weeks, and even then the final price often comes down to gut feel.

Increase pWin, protect margin, and give your team weeks back. - Revere, company site

That last line hints at the wedge. Revere is not trying to write the whole proposal. It is going after the single decision that most shapes the outcome and is hardest to get right: what to charge, and why.

3Agent stages: research, model, simulate
pWinThe number everything reduces to
F25Y Combinator batch

The ProductFrom Fragments to a Probability

Revere's platform moves through three stages, each handled by agents rather than analysts. The output is not a report to read but a number to act on.

How it works
One pursuit, three agent stages
01
Market & Cost Research
Agents gather labor-rate data, job postings, other-direct-cost analysis, and comparable awards to ground the bid in real numbers.
02
Competitor Modeling
Builds bidder profiles and estimates each rival's wrap rate and likely price, turning guesswork into a range.
03
Competitive Simulation
Runs Monte Carlo simulations across scenarios to produce a modeled pWin at each price point you might submit.
The pipeline reads left to right, but the value lands at the end - a probability you can defend in a pricing review.

The Monte Carlo step is worth pausing on. It is a technique borrowed from physics and finance, where you run the same scenario thousands of times with slightly different assumptions to see the shape of the outcomes. Applied to a bid, it answers a question no spreadsheet cleanly can: given everything we can infer about who else is at the table, how often do we win if we price here?

Illustrative model
Modeled pWin climbs as price drops
$ High
40%
$ Mid
60%
$ Low
80%
A stylized version of Revere's core output. Lower prices lift win probability but eat margin - the model makes that trade-off visible instead of intuited. Figures are illustrative, not company data.

Under the HoodReading a Rival's Rate Card

Much of the work is reverse-engineering cost. A competitor's price is built on labor categories - a program manager here, a systems engineer there - each with a rate, marked up by an overhead "wrap." Revere estimates those inputs from public signals. The sample rate ranges it publishes give a feel for the resolution it works at.

Sample labor categories
The building blocks of a bid
Labor CategoryModeled Hourly Range
Program Manager$72 - $96
Senior Analyst$48 - $66
Systems Engineer III$61 - $86
Sample ranges from Revere's site. Stack enough of these, apply a wrap rate, and you have an estimate of what a rival will charge.

Stack those categories, apply an estimated wrap rate, and a competitor's opaque bid starts to resolve into a range. Do it across the likely field, feed it into the simulation, and the pricing decision stops being a leap of faith.

The DifferenceAn Agent, Not a Dashboard

Plenty of software promises government contractors better data. What separates Revere's approach is where the analysis happens. The old model hands you a dashboard and leaves the interpretation to you. Revere's bet is that the interpretation is the product.

The manual way

  • Weeks assembling scattered intel
  • Competitor pricing is a guess
  • Final number set by gut feel
  • Hard to defend in a pricing review

The Revere way

  • Agents compile the field in hours
  • Rivals' rates modeled from signals
  • A pWin attached to each price
  • A rationale you can point to

There is a quieter advantage too. Revere says each client pursuit feeds a proprietary competitive database that grows with use. The more bids the platform models, the sharper its picture of the field becomes - the kind of compounding edge that is hard for a newcomer to copy and hard for a customer to walk away from.

Your unfair advantage. - Revere

The MarketWhere Revere Fits

Revere lands in a young category of AI tools built specifically for federal contracting, alongside platforms such as GovDash that aim to modernize how contractors pursue and manage work. But Revere's real competition is the status quo: the spreadsheets, the manual research, and the instinct-based pricing that most teams still rely on. Its focus is narrower than a full proposal suite and, for that reason, sharper - it owns the pricing decision and tries to own it completely.

The founding team is small - Drew Taylor serves as co-founder and chief executive, working alongside co-founders Matthew Fan and Shreyas Chennamaraja out of San Francisco. That is a lean crew to take on one of the least transparent corners of the economy, which is roughly the point. The unglamorous verticals tend to have the least competition and customers who feel the pain most acutely.

Whether Revere becomes the pricing brain for government contracting or one option among several, its wager is clear enough: in a market where the winning move is a number, the company that models that number best has, well, an unfair advantage.

#govcon#ai-agents#pricing-intelligence #pwin#competitive-intelligence#federal-contracting #yc-f25#monte-carlo#saas