Engineer and entrepreneur

I build AI systems that sell

Chat assistants, CRMs, Telegram bots and mobile apps. Everything below is already live — with users, payments and history in the database. Working products, not concepts.

Case studies

Six systems in production

The numbers come straight from the database and were updated today. On a call I'll show you any of these systems from the inside: code, admin panel, metrics.

Mobile · Legal-tech in two app stores

Musofir — a mobile app with a legal assistant

Helps migrants from Central Asia deal with paperwork. Inside is not a plain text bot but an assistant with a rich interface: it reads a passport from a photo, checks employers and entry bans, shows government offices on a map, sets deadline reminders, searches flights and answers with images. Plus consultations and a support call — right from the app.

I built it end to end myself — from Flutter and the backend to a dozen external APIs, SBP instant payments and releases in two app stores.

4 languagesRU · UZ · TJ · KG
2 storesGoogle Play · RuStore
FlutterNestJSPostgreSQLLLMrich chatdocument checksmaps · flightsSBP paymentsin-app calls
14:02
Personal lawyeronline
Message…
Home Lawyer Services Profile

AI · Sales

A front-line assistant: from "hello" to a lead in the CRM

Answers on the website and in Telegram around the clock: it knows the company's pricing and services, remembers past conversations and asks clarifying questions. When the client is ready, it logs the lead in the CRM and wakes the manager with a notification.

131,407messages handled
34,582conversations
~₽10/mocost per client
LLMknowledge baselong-term memoryCRMTelegram
client-site.ru + CRM

CRM · Legal services · AI used daily

A law firm CRM with a court document generator

Tracks leads and deals around one specific law firm's workflow. The core feature is a document generator built on reference templates: claims, objections, complaints, motions and acts come together in seconds. The lawyer proofreads a draft instead of writing it from scratch for an hour.

dozensof documents generated daily
~15 secper draft instead of an hour by hand
₽0/moin SaaS subscription fees
ReactNestJSPostgreSQLLLMreference libraryDOCX
Open the live system
sovetnik-crm.24experiment.ru

In-app voice · WebRTC

The client calls from the app — the operator answers in the browser

This is not telephony: the call goes over the internet, like a voice call in a messenger — no phone numbers, SIM cards or per-minute charges. The client taps "call" in the mobile app, and the operator picks up right inside the client's card. Built from scratch on WebRTC — with IVR, call recording, inbound and outbound.

1,466in-app calls
8,773support messages
₽0for connectivity and licenses
WebRTC P2PWebSocket signalingTURNcall recording
crm.pravo-centr.ru/leads

B2B · White-label

An embeddable AI chat for partners

A partner adds one line of code — and an assistant under their own brand appears on their site. Conversations and leads flow into a shared CRM, and updates reach every partner at once.

1 lineof code on the partner's site
~₽0.5cost per answer

Telegram · Automation

Telegram bots and mini apps

Booking, payments and a service catalog without leaving Telegram. Runs on its own backend instead of bot builders — no monthly fees, no subscriber limits.

−100%bot-builder subscription fees
your datathe business owns the entire database

Services

Three ways to work together

Fixed timeline and price — no "we'll estimate after discovery". The first half hour is a free breakdown of your task.

most popular

AI sales assistant

A front line on your website and in Telegram that consults, qualifies and turns visitors into leads.

  • Trained on your knowledge base and price list
  • Leads go straight to the CRM, managers get notified
  • Client memory and conversation analytics
from ₽150,0002–3 weeks
Discuss your project

Automation and CRM

I remove manual routine from your processes: a system built for your business instead of spreadsheets and chat threads.

  • A CRM shaped to your processes, not the other way around
  • Bots, integrations, documents, reports
  • Calls and messaging in one window
from ₽200,0003+ weeks
Discuss your project

Turnkey product

An MVP all the way to the app stores: mobile app, backend, payments, analytics. One contractor instead of a team.

  • From idea to app-store release
  • Subscriptions, payments, user accounts
  • Self-hosted infrastructure — several times cheaper than the cloud
from ₽500,0006+ weeks
Discuss your project

I build the prototype before any contract or prepayment. If it's not a fit, you owe nothing.

Approach

Why it's faster with me

Prototype first, contract second

In three to five days you have a clickable prototype running on your data. You decide on the contract after you've tried it with your own hands.

I speak the language of business

Ten years in management roles in telecom and legal services. We talk leads, conversion and money — not story points.

One engineer — the whole stack

Mobile, backend, AI, payments, servers. Any question gets resolved in one conversation, not through a manager and three subcontractors.

I show working systems, not slides

Every case study can be opened and tried: the app is in the stores, the CRM is in daily use, and the assistant is right on this page.

Process

From task to production in four steps

Task breakdown

A 30-minute call: what hurts, where money is leaking, and what we automate first.

30 minutes · free

Prototype

A working prototype on your data — not a Figma mockup, but a system you can click through.

3–5 days

Delivery

Short iterations with a demo every week. You see progress as it happens — no surprises at the end.

2–6 weeks

Support

After launch I monitor the system and keep improving it. If something breaks, I fix it — not you.

by subscription

About me

JZ
Jahongir Zairov

An engineer with a management background

Over ten years in management roles in telecom and legal services: sales, call centers, operations, hiring. I saw from the inside how businesses lose clients on the front line — and started building systems that fix it.

Today I have my own product on Google Play and RuStore, plus CRMs and assistants running for clients. With clients I talk business first: economics before technology.

Management roles: telecom, legal services, customer service
Stack: Flutter, React, NestJS, PostgreSQL, LLM integrations, WebRTC, payments
Own product: Musofir — a legal-tech app on Google Play and RuStore

Tell me about your task — I'll show you how AI solves it

A half-hour breakdown is free. If I see the task isn't for me, I'll say so right away.

Or message me directly

Telegram@zja_77 Email