Nikolay Baakh

• — Built a natural-language-to-SQL/Python agent over Postgres/ClickHouse with schema introspection and execution safety nets; scales to additional MCP connections as the service grows
• — Implemented an auto-updating pgvector RAG knowledge base with hybrid retrieval, re-embedding, access control, and per-answer confidence; added deep research and sandboxed code execution for charts and calculations
• — Built a React + TypeScript admin UI with observability, logging, documentation, and runtime settings; shipped Slack bot and Jira integration
• — Productionised end-to-end: streaming chat, OIDC/Keycloak SSO with RBAC, observability, Docker + GitLab CI deploy; designed and documented the full API through Swagger
• — Led security & reliability hardening across ~50 tested fixes — SSRF, IDOR, OIDC account-takeover, concurrency bugs

• — Built an AI usage framework for analysts, cutting task execution time across the team
• — Built and led a 10-person data team (2 engineers, 4 analysts, 2 full-stack devs, 1 PM, 1 web analyst)
• — Designed an all-inclusive data strategy for the B2B SaaS firm; applied advanced analytics for customer engagement and KPI tracking
• — Developed ClickHouse data warehouse for real-time insights; engineered user behavior tracking via PostHog
• — Deployed Apache Superset for real-time product metrics, platform analytics, and B2B CLV dashboards
• — Built robust data pipelines and APIs for frontend-backend data delivery interfaces

• — Developed an AI-driven framework for the data engineering team, roughly halving pipeline shipping time
• — Built data department from scratch; hired and directed a team of 5 (2 engineers, 3 analysts)
• — Architected scalable ELT pipeline (Python + dbt + Airflow → ClickHouse), integrating Affilka and Affise APIs
• — Built buyer-side traffic payback model, enabling traffic valuation and distribution decisions
• — Deployed Holistic for company's first centralized dashboards; surfaced key structural financial issues for the executive team

• — Introduced AI tooling into the data analytics workflow across the org
• — Coordinated 3 full-cycle analytics teams across Data, Engineering, BI, Web, and Business Analytics
• — Executed large-scale ETL refactoring; led strategic planning, resource plan, and department budget
• — Integrated user retention costs (bonuses, promo codes, call center, support) into traffic payback formulas
• — Drove a product-centric approach with the business; tightened quality control around business value and customer satisfaction
• — Created competency matrices and individual development plans for team leads

• — Established and launched a Marketing Analytics team of 6 Data Analysts from scratch
• — Implemented data analysis standards for the company's marketing activities
• — Built key BI tools the leadership relies on for operations: end-to-end real-time traffic analytics, employee efficiency, traffic payback, and user cohort analysis
• — Introduced AOV-based ROAS and ROI analysis method for the marketing department
• — Mentored analysts: Junior → Middle+ and Middle → Senior levels

• — Designed custom data extraction pipelines integrating third-party APIs with cleaning and transformation stages
• — Built interactive Looker dashboards for key stakeholders, enabling quick access to critical metrics
• — Maintained and expanded databases with 2,000+ manual records per month, ensuring data quality and consistency

Gained access to Postgres-based 1C and CRM databases under NDA to perform data analysis for a leading 26-store finishing materials retailer in St Petersburg. Discovered the need for data analysis and self-initiated the role while employed as Sales Manager.

Developed and executed social media strategies for multiple clients. Began collecting and analysing campaign performance data, which led to the discovery of Data Analysis as a career path and a self-directed transition into the field.