For AI teams facing security questionnaires
When the questionnaire slows a deal, see which AI answers your public record can support, which points need team evidence, and what to fix first before the review expands.
Published operating record
Evidence about Walseth AI's own work sample, not a promise about your repo.
No private access required for the scan. No compliance certificate implied. No paid work starts until the business pressure and scope are clear.
Saved output example: vllm-project/vllm scored 78/100, grade B, on Mar 16, 2026. Use the live demo for current state.
Public evidence first. Unsupported points stay visible before paid work.
Security review packet preview
Sample public-scan output, not customer data
Buyer question
AI use, data flow, retention, and ownership
Captured
Public answer
What the repo, policies, and routes support today
Mapped
Next fix
What to address before the next buyer email
Ranked
Model use
Visible now
Public files show where automation, scanning, or report generation appears.
Still unsupported
Confirm owner, model purpose, and customer-facing limit before the answer is final.
Data handling
Visible now
Policies, API routes, and page copy show what the public record can support.
Still unsupported
Add a team-owned retention answer if procurement asks for private-system detail.
Review controls
Visible now
Tests and route boundaries show where proof, action, and caveats stay separate.
Still unsupported
Name the escalation path for answers that need private evidence.
Example answer map
Bring the first defensible answer before the second buyer email.
The scan starts from public evidence, keeps unsupported points visible, and turns the next paid step into a narrow review packet.
Product you can inspect
Walseth's demo turns a public GitHub project into a score, sample findings, and a narrow next step while keeping private system details outside the scan.
Saved demo-output snapshot
Saved Mar 16, 2026. Use the live demo for current state.
78
Grade B
Git hooks configured
7 CI/CD workflow(s)
Security policy present
This shows the output format: score, grade, public findings, and timestamp. It does not certify private controls or changes after the saved scan.
01
Paste a public GitHub project. The demo reads visible product signals only, before any private access.
02
The result shows a score, sample findings, and a suggested next move without turning unknowns into finished answers.
03
The record shows how Walseth reports its own measured work before a team asks for review.
Built around buyer questions
The point is not to make the repo sound safer than it is. The point is to make the visible answer clear enough that your team knows what to send, what to fix, and what still needs private evidence.
The scan reads public signals: AI use, data-flow notes, policy language, tests, owners, and release context.
Walseth AI separates defensible answers from points that still need team review or private-system evidence.
The Baseline Sprint turns a real deadline into ranked fixes, answer language, and buyer-ready starter materials.
The path stays deliberately narrow
This keeps the first move useful without turning every visitor into a sales conversation.
01 / Free first read
Use the scan to see the answers a reviewer can infer before your team grants private access or starts a sales process.
Run Free Repo Scan02 / Fit check
Baseline review checks the buyer question, deadline, repo, and scope so paid work starts only when the gap is specific.
Request Baseline Review03 / Paid sprint
When the gap is worth fixing, the sprint produces the AI inventory, answer map, ranked gaps, and starter materials.
See PricingWhen the gap is worth fixing
A fixed-scope sprint for one AI product or repo family: map the AI use and data flow, identify the review gaps most likely to slow a deal, and produce buyer-ready materials your team can use immediately.
What the sprint produces
Monitoring and retained execution come later only when there is a baseline worth maintaining.
Request Baseline SprintUseful before the meeting
Inspect how Walseth AI reports its own work before asking it to review yours.
See the kinds of public-repo gaps that can slow a review conversation.
Use this when security review, procurement, audit, or launch timing is already active.
Ready to start
The free scan gives the first read. Baseline review is for moments when a sale, audit, or launch needs a stronger answer soon.