AI Engineer
Manila · UTC+8 · remoteIvan Kharlamov
I build systems for marketing and advertising: data infrastructure, automation and LLM-based agents. Previously at Yandex — I know advertising from the inside, from buying traffic to deals in the CRM.
01What I do
Data infrastructure
Collecting data from ad platforms and CRMs, storing it in PostgreSQL, scheduled daily ETL, reconciliation and data quality checks.
Automation
Cloud functions that handle the routine without a human: rebuilding product feeds, computing reports, tracking changes and reporting them.
LLM agents
Applied agents for real work: keyword research and cleanup, text and image generation, MCP integrations, and cost control per request.
02Stack
Languages and data
- Python
- SQL
- PostgreSQL
- ClickHouse
- TypeScript
- pandas
Pipelines and infrastructure
- Airflow
- Cloud Functions
- Docker
- S3
- Git
- CI/CD
- Vercel
Models and agents
- OpenAI API
- Anthropic API
- Gemini API
- OpenRouter
- MCP
- RAG
- Vector databases
- LLM output evaluation
Analytics and BI
- Yandex DataLens
- Power BI
- Yandex Metrica
- Google Analytics
- End-to-end attribution
Ad and CRM APIs
- Yandex Direct API
- Google Ads API
- Google Analytics API
- Yandex Metrica API
- Bitrix24 API
- Telegram Bot API
- Wordstat API
03Experience
Late 2025 — present
AI engineer
Education company · internal development
- Building LLM-based systems for real advertising work: keyword research and cleanup, ad copy and image generation, data analysis.
- Assembled a multi-agent environment with reusable skills and external data sources connected over MCP.
- Moved manual operations — feed assembly, reporting, campaign preparation — into automated cloud processes.
- Track model cost and output quality so the automation stays predictable to run.
- Side project in parallel: automation and analytics for a medical education centre.
2024 — late 2025
Performance marketer, then data engineer
Same company
- Owned traffic buying in Yandex Direct and Google Ads, including international markets.
- Built end-to-end analytics: ETL from ad platforms and CRM into PostgreSQL, linking spend to closed deals.
- Dashboards and recurring reporting, plan-versus-actual by channel.
- Gradually moved from running campaigns by hand to building the tooling behind them.
2023 — 2024
Senior advertising strategist
Yandex
- Worked with major advertisers and budgets upwards of 100M RUB per month.
- Auction mechanics, algorithms and product internals of Yandex Direct from the platform side.
- Scaling strategies and campaign performance reviews.
04Projects
→Product feed robot
Rebuilds a product feed for ad platforms every day and tracks catalogue changes.
Python · Yandex Cloud Functions · PostgreSQL
End-to-end analytics
Links ad spend to leads and deals from the CRM to measure cost per lead and cost per sale.
Python · PostgreSQL · ETL
Agent system for marketing work
A set of agents and skills that carry repetitive advertising work end to end.
LLM · MCP · Python
05Certificates
Professional certifications for ad platforms. Not strictly engineering, but they exist.
