You will get an AI agent MVP: GPT/Claude assistant integrated with your business tools
Project details
You will get a working AI agent MVP: a GPT or Claude assistant that does real work inside your tools. It answers customers, updates your CRM, drafts emails and documents, books meetings, queries your data.
I build AI products end-to-end (see my portfolio: a personal finance platform with four clients on one backend, production Telegram bots, n8n automations), so your agent ships with the unglamorous parts that make it reliable: function calling done right, guardrails, error handling, logging.
Former e-commerce founder: I scope agents around business outcomes, not demos. You get an architecture plan before I write code, weekly iterations on real scenarios, and full ownership of the result. Message me with your use case and I will reply with a concrete plan within hours.
I build AI products end-to-end (see my portfolio: a personal finance platform with four clients on one backend, production Telegram bots, n8n automations), so your agent ships with the unglamorous parts that make it reliable: function calling done right, guardrails, error handling, logging.
Former e-commerce founder: I scope agents around business outcomes, not demos. You get an architecture plan before I write code, weekly iterations on real scenarios, and full ownership of the result. Message me with your use case and I will reply with a concrete plan within hours.
AI Algorithms
Large Language Model, Multimodal Large Language Model, Transformer ModelAI Applications
AI Chatbot, Conversational AIAI Development Language
PythonAI Models
ChatGPT, GPT-4, WhisperWhat's included
| Service Tiers |
Starter
$400
|
Standard
$2,400
|
Advanced
$3,900
|
|---|---|---|---|
| Delivery Time | 7 days | 21 days | 30 days |
Number of Revisions | 1 | 2 | 3 |
AI Model Integration | |||
Batch Normalization | - | - | - |
Database Integration | |||
Detailed Code Comments | |||
Image Upscaling | - | - | - |
MLOps | - | - | - |
Model Deployment | - | ||
Model Documentation | |||
Model Monitoring | - | - | |
Model Testing & Optimization | - | ||
Model Tuning | - | - | - |
Natural Language Processing | - | - | - |
NLP Tokenization | - | - | - |
Pre-Training | - | - | - |
Prompt Engineering | |||
Setup File | |||
Source Code |
Frequently asked questions
About Valeriy
AI Automation Engineer | Chatbots, Workflows & AI Agents for Business
Astana, Kazakhstan - 10:33 pm local time
What I've built:
AI Personal Finance Platform — one product, four clients on a shared backend: Telegram bot, Telegram Mini App, web app, and iOS app (App Store release: August 2026). It accepts bank statements, screenshots, and voice messages, runs LLM analysis, and surfaces insights, budgets, and trends. Stack: Python, FastAPI, PostgreSQL, OpenAI/Claude, Whisper, Telegram Bot API, React + TypeScript + Tailwind + Framer Motion.
AI English Tutor (Telegram) — adaptive learning system with automated CEFR level assessment, AI-generated daily lessons, voice practice, and gamified progress tracking. Duolingo-style experience, fully inside Telegram.
Algorithmic Trading R&D (personal project) — built and ran an automated stock/ETF strategy on Interactive Brokers API to explore how far LLM-driven news analysis can go in financial decision-making. Two-basket system with VIX-adaptive risk controls and Telegram monitoring interface. Included as a demonstration of complex API integration and AI-driven decision systems — not a client service I offer.
Tech stack: Python, FastAPI, PostgreSQL, Supabase, OpenAI API, Anthropic Claude API, Whisper, Telegram Bot API, React, TypeScript, Tailwind, Framer Motion, Docker, nginx, Cloudflare Workers.
What sets me apart: I'm not a developer who learned about business — I'm a former business owner who learned to build. I ran my own clothing brand for over three years: sourcing from China, managing cross-border operations between Russia and Kazakhstan, scaling on the Wildberries marketplace, and hiring an operations team (30+ interviews from a pool of 200+ candidates). Before that, I ran an offline-first clothing brand with a physical store and Instagram-driven customer acquisition. I've also traded crypto arbitrage hands-on — giving me real domain understanding in financial automation.
When I build automation for you, I think in unit economics and operational reality. The question isn't "does the code work" — it's "does this move a real business metric."
Steps for completing your project
After purchasing the project, send requirements so Valeriy can start the project.
Delivery time starts when Valeriy receives requirements from you.
Valeriy works on your project following the steps below.
Revisions may occur after the delivery date.
Scope & architecture
We define the agent's tasks, tools and success criteria. You get an architecture plan before development starts.
Build & iterate
I build the agent, connect your tools and we test it on real scenarios with weekly check-ins.