Audit of an existing codebase and architecture by Sevastian RakhimovAudit of an existing codebase and architecture by Sevastian Rakhimov
Audit of an existing codebase and architectureSevastian Rakhimov

Who this is for

You are about to spend real money on a codebase and you cannot read it yourself.
Maybe an agency delivered something and you want a second opinion before the final payment. Maybe your developer left and you need to know what you actually own. Maybe an AI wrote most of it and it works, which is exactly what worries you. Maybe you are about to acquire a product and the pitch is better than the repository.

What you get

A written report, in plain language, that answers four questions:
What is actually there. The stack, the architecture, how the pieces talk to each other, what is custom and what is a library, and how much of it is genuinely finished versus scaffolded.
What will break first, ranked. Not a list of style complaints — a list of things that will cost you money or users, ordered by how soon. Security holes, data model decisions that get expensive later, missing error handling, single points of failure, and the parts that only work because one person remembers a manual step.
What it will cost to fix. Rough effort per item, so you can decide what to do now, what to do later and what to live with. Some of it you can hand to your existing developer.
Whether the AI part is real. If there is an LLM in the product: is it measured, is it traced, does it fail safely, and what does it cost to run at your volume. AI features that impress in a demo and degrade quietly in production are the single most common thing I find.
You get the report whether or not you ever hire me for the fixes, and it is written so another engineer can act on it without me.

Why this is worth $500

Because I have been on both sides of it. I built and still run F/AI — an AI fitness trainer with 340,000+ users, $8,000 MRR and billing across three payment sources — and before that spent six years with a New York SaaS company, the first eighteen months as its only engineer.
So this is not a linter with an opinion. It is a judgment about what will hurt you, from someone who has been paged at 3am by their own mistakes.

How it works

Send me read access to the repository and a couple of sentences about what the product is meant to do. If there is something specific worrying you, say so and I will look at it first. You get the report as a document, and a call to go through it if you want one.
FAQs

Starting at$500
Tags
Flutter
AI Engineer
Backend Engineer
DevOps Engineer
Fullstack Engineer
LLM
Mobile Engineer
Software Engineer
Code Review
Technical Due Diligence
Service provided by
Sevastian Rakhimov Tbilisi, Georgia
2
Followers
Audit of an existing codebase and architectureSevastian Rakhimov
Starting at$500
Tags
Flutter
AI Engineer
Backend Engineer
DevOps Engineer
Fullstack Engineer
LLM
Mobile Engineer
Software Engineer
Code Review
Technical Due Diligence

Who this is for

You are about to spend real money on a codebase and you cannot read it yourself.
Maybe an agency delivered something and you want a second opinion before the final payment. Maybe your developer left and you need to know what you actually own. Maybe an AI wrote most of it and it works, which is exactly what worries you. Maybe you are about to acquire a product and the pitch is better than the repository.

What you get

A written report, in plain language, that answers four questions:
What is actually there. The stack, the architecture, how the pieces talk to each other, what is custom and what is a library, and how much of it is genuinely finished versus scaffolded.
What will break first, ranked. Not a list of style complaints — a list of things that will cost you money or users, ordered by how soon. Security holes, data model decisions that get expensive later, missing error handling, single points of failure, and the parts that only work because one person remembers a manual step.
What it will cost to fix. Rough effort per item, so you can decide what to do now, what to do later and what to live with. Some of it you can hand to your existing developer.
Whether the AI part is real. If there is an LLM in the product: is it measured, is it traced, does it fail safely, and what does it cost to run at your volume. AI features that impress in a demo and degrade quietly in production are the single most common thing I find.
You get the report whether or not you ever hire me for the fixes, and it is written so another engineer can act on it without me.

Why this is worth $500

Because I have been on both sides of it. I built and still run F/AI — an AI fitness trainer with 340,000+ users, $8,000 MRR and billing across three payment sources — and before that spent six years with a New York SaaS company, the first eighteen months as its only engineer.
So this is not a linter with an opinion. It is a judgment about what will hurt you, from someone who has been paged at 3am by their own mistakes.

How it works

Send me read access to the repository and a couple of sentences about what the product is meant to do. If there is something specific worrying you, say so and I will look at it first. You get the report as a document, and a call to go through it if you want one.
FAQs

$500