You built it yourself with Claude or Cursor. You need someone to tell you if it's right.
A written assessment of your AI system or plan: whether it will work, how it fails, what it costs to run and what to fix first. We make the three highest-risk fixes ourselves.
You have this problem if…
- An AI coding tool wrote most of the code, and nobody has read all of it.
- You're about to pay for a build and want a second opinion on the design first.
- An investor, customer or security reviewer asked about permissions, audit logs, failure handling or running cost, and you didn't have answers.
What you get
- A written assessment covering feasibility, failure handling, permissions, audit logging, running cost and security
- A map of the business workflows in your code and where each one lives
- Risks ranked by impact, each with evidence from your code
- A proof-of-concept plan with acceptance criteria, if you are still at the plan stage
- The three highest-risk fixes, made in your repository
How it is done
- 01
Kickoff
A call to understand what the system is for, who uses it and what worries you.
- 02
Map the code
UnvibeCode traces the business workflows, the connections between files, and where behavior actually lives.
- 03
Review
Prompts, tools and permissions, data flow, failure handling, cost per request and security, each checked against the code.
- 04
Fix the top three
The three highest-risk issues are fixed in a branch you can review.
- 05
Walkthrough
The written memo plus a 45-minute walkthrough call.
What usually goes wrong, and how we handle it
Generated code hides the workflow
When a model wrote the code, the business logic is spread across files nobody chose. You can't review the design until the workflow is mapped.
Agents with broad permissions
An agent that can write to production, send email or spend money needs limits and an approval step. These are the most common high-risk findings.
Cost per request is rarely measured
A system that works at ten users can lose money at a thousand. We estimate cost per request from the code paths.
Why you can trust us with this
UnvibeCode, the open-source tool our founder built, reverse-engineers a codebase into business workflows, a connected-code map and evidence-backed risk findings. Sample analyses of four public repositories are online.
Our founder reviews manuscripts in applied AI for Elsevier and Springer Nature journals, and was a program-committee reviewer for the NeurIPS 2026 Verify-Agents workshop.
What we don't do on this package
- No rubber stamps. If the design won't hold up, the memo says so.
- No rewrite as a condition of the review.
Questions
Do you need production access?
No. Read access to the repository is enough. Production access only if you want deployed configuration checked.
Can you review a plan before anything is built?
Yes. You get the same assessment against the design, plus a proof-of-concept plan with acceptance criteria.
Who does the work?
Suresh Rajendran, our founder, does the engineering on every engagement himself. Nothing is handed to a junior engineer or a subcontractor.
Where does our data go?
We work inside your accounts: your cloud, your repository, your model-vendor keys. Where a vendor offers zero data retention we turn it on, and your data is never used to train a model. Our access is removed and credentials are rotated at handover.
How do we pay?
The first step is paid up front. Builds are 50% to start and 50% on acceptance. We invoice from Destello AI Incorporated in US dollars, payable by ACH or card. If you go ahead with a build within 30 days of the first step, its fee is credited.
What if the result doesn't meet the bar?
Acceptance criteria are agreed in writing before work starts, and the second payment is due only when they're met.
Will you sign our NDA?
Yes, usually within one business day.
Get an independent review this week.
$1,200–$1,500, credited to the build if you go ahead within 30 days. Book a 20-minute call, and you'll have a written scope with acceptance criteria within one business day.
Or email hello@destello.us or call (628) 202-5322. We reply within one business day.