Before we've had a single call with a new client, Pulse tells us exactly what's broken, what's missing, and what it will cost to fix. That's how we walk into every discovery call with the diagnosis already done.
Health score
acme-saas.vercel.app
AI gap analysis
3 critical gaps · 8 build opportunities · 3-phase roadmap
Try it live — free
The same engine, no AI, no signup — your health score in seconds. This is the exact widget you can drop on your own site.
Add it to your site
<script async src="https://foundry.gitwork.co.uk/embed/pulse/embed.js"></script>The opportunity
Every week, developers and founders use Bolt, Lovable, Cursor, and other AI tools to ship products they couldn't build before. The code runs. The product works. But it's nowhere near ready to handle real users, real money, or real security requirements.
That gap — between "it works on my machine" and "production-ready" — is Gitwork's market. Pulse is the tool that lets us quantify it, prove it, and sell the fix.
Before Pulse
We walk into discovery calls and spend 45 minutes asking the client what they've built. We leave without a clear scope. The proposal takes days.
After Pulse
We scan the product before the call. We walk in with a health report, three critical gaps highlighted, and a scoped engagement ready to sign.
Works on everything
Paste a URL, an App Store link, a GitHub repo, or just describe what the client built. Pulse figures out the rest.
The output
A full report — automated checks plus AI analysis — ready before you pick up the phone.
A single number that anchors the conversation. We can benchmark it against previous scans and track improvement over an engagement.
Specific things that are broken, missing, or dangerous — with urgency ratings (CRITICAL / HIGH / MEDIUM) and business impact for each.
Every gap is a billable opportunity. Each one comes with an effort estimate (S / M / L / XL) and business value rating — ready to scope.
A phased plan — Stabilise, Productionise, Scale — with goals and timelines. Becomes the structure of the proposal.
12–20 items covering legal, auth, payments, onboarding, trust, and observability. DONE / PARTIAL / MISSING status for each.
One click. A clean, professional PDF of the full scan. Print it, share it, or use it to open the call.
Automated checks
Pulse doesn't run generic checks. It detects the project type first — AI app, SaaS, mobile app, marketing site, API — then runs only the checks that matter. Purpose-built for 2026: it knows what an LLM product needs in production, and what AI code generators leave behind.
AI analysis
Switch AI providers in Settings. No code changes needed.
If no key is configured, Pulse still runs all 500+ automated checks and returns a structured mock analysis — so the whole workflow can be tested without spending API credits.
Claude
✓ READYclaude-opus-4-8
Default. Best for structured gap analysis.
GPT-4o (OpenAI)
✓ READYgpt-4o
Any OpenAI model. Model name is configurable.
Gemini (Google)
✓ READYgemini-2.0-flash
Via OpenAI-compatible endpoint. No extra SDK needed.
Local LLM
✓ READYOllama / LM Studio
Point at any OpenAI-compatible server. Zero API cost.
How to use it
Paste the client's URL, GitHub repo, or App Store link into Pulse. The scan runs in 30 seconds. You now have a health score, critical gaps, and a scoped set of build opportunities before you've spoken to them.
Open with the health score. "We scanned your product and it's sitting at 58/100 — here's what that means." Walk them through the top gaps. The client didn't know half of these existed. You've already demonstrated more value than any other vendor they'll talk to.
Download the PDF report. Send it to the client as your follow-up. The report becomes the basis of the proposal — scope, phases, and priorities are already defined by the scan.
Paste any URL and get a full health report in under 60 seconds. No API key needed to test.