# Marcus Delgado, VP of Business Development at Ridgeline Civil Group — read of Contract Bidding AI, May 22, 2026

> 14 years in federal contracting, currently running BD for a 47-person civil infrastructure firm that bids mostly USACE and DOT work. Coaching my kid's Little League takes every Saturday morning I might otherwise spend cleaning up proposal drafts.

## How I got here

Google search. I typed something like "RFP proposal drafting software small contractor" because we lost a $2.1M bridge inspection contract last quarter and our capture manager thinks we're under-bidding on volume. LinkedIn has been showing me AI proposal tools for six months straight and I've ignored all of them. This one came up organically, not as an ad, which is why I clicked.

## What I clicked first

The header does its job: "Win more government contracts without the $5K-15K per bid consultant cost." That number is real. We paid $11,200 to a BD consultant on our last IDIQ and came in third. So the hook landed.

Then I kept reading and hit: "You write one proposal every quarter and hope it lands." That felt like someone had looked at my calendar. We bid maybe five proposals a year. That line made me lean in a little.

## Where I paused

The stat block. "75% faster bid drafting. $50K saved per bid vs. consultants. 40% higher win rate for repeat bidders." I stopped here for a full minute. Not because I believed it. Because I was trying to figure out where those numbers came from. There's no source, no "based on X customers over Y bids," nothing. They're just sitting there like facts. The 40% win rate claim in particular -- that's a career-defining improvement if real. Which means if it were real, there would be case studies, a customer quote, a named firm somewhere. There's none of that.

## What I distrusted

This part: "Honest disclosure: we don't have live customers on this idea yet."

That's buried below the fold after three screens of marketing copy. And it's jarring because the entire page talks about the product as if it exists and works. "Contract Bidding AI analyzes the RFP, your past wins..." reads like a product that's running. But it's not. It's a strategy package and some code starter. The compliance automation feature -- CAGE code, DUNS, past performance auto-population -- that's not a content problem, that's integration work. I've seen firms spend six figures building that out. The page says it like it's a checkbox.

Also this: "pain intensity: 4/10." That's their own score for their own product's pain intensity. If the people selling this rate the pain at a 4, I don't know what to do with that.

## What would convince me

One real proposal. Redacted, fine, but show me an RFP that came in, show me what the AI produced for the technical approach and management sections, and show me the scoring feedback from the agency if the bid was debriefed. Even a losing bid with a debrief score of 78/100 would tell me this thing produces something usable. Screenshots of draft output would help too. I don't need a win. I need to see the work product.

A named firm that used this, even in beta, even for free. "We piloted with a 12-person IT services firm in Reston" -- that's enough to continue the conversation.

## What I'd ask in an email reply

1. The page mentions "your firm's past wins" as an input to the system. How does that actually work? Do I upload past proposals manually, connect to some database, or describe them in a form?

2. CAGE code and past performance auto-population -- is that live integration with SAM.gov, or does someone on my team configure a profile once and it pulls from that? Because the difference between those two things is months of work.

3. The "$50K saved per bid vs. consultants" -- is that modeled from Fermi math or from actual contractor data? If it's modeled, what are the assumptions?

## Verdict: on-the-fence

The pain is real, the hook is credible, and whoever wrote the copy has clearly talked to someone who actually bids government contracts. But "no live customers" plus stat claims with no sourcing plus a product that might not exist yet puts me in wait-and-see. I'd bookmark it and check back in six months.

---
*Memo by skeptic persona, generated 2026-05-22. Studio breaks own self-grading loop.*