Anastasia Bogacheva — analytics engineer. I build analytical systems for money that has to be counted correctly — payments, risk, fraud, compliance. When the tooling doesn’t exist, I build it myself.
// IF IT HAPPENS TWICE, AUTOMATE IT.
Fraud exposure needed a policy answer: how much outbound volume should a customer's trusted inbound activity actually earn? I built the eligibility analytics — simulated four policy thresholds, quantified the volume at risk, and shipped the Power BI views risk leadership steers by, with violation reporting by country and segment.
A risk-scoring pipeline the business already trusted had to move engines without the numbers moving. I migrated the 14-model chain from SQL Server to BigQuery and validated it to row-level parity — including root-causing a silent cross-engine mismatch to nondeterministic ROW_NUMBER ordering, then pinning it and proving the match.
New messaging-session data had to become queryable without becoming expensive. I built the dbt staging and Airflow orchestration for a 147-million-row, 18 GB table, partitioned so queries stopped paying for it — then ran a warehouse cost hunt through INFORMATION_SCHEMA that turned partitioning and clustering into visibly smaller bills, with analyst workflows unchanged.
EMPLOYER WORK DESCRIBED AT A SAFE ALTITUDE — FULL DETAIL IN CONVERSATION.
A job-search intelligence framework — part of a larger AI career operating system built in Obsidian. Listings flow in, get scored against my profile, and land in Telegram, with an interview-prep engine that builds a briefing for every callback.
An eight-skill analysis suite for Claude Code: business context → exploration → methods → statistics → SQL → visualization → reporting, with an orchestrator that sequences the whole workflow.
A personal tracking system with RPG mechanics — habits, energy and streaks become quests and stats. Same pipeline architecture as JobSeek, pointed inward: tracked in Obsidian, rendered as generated dashboards.
If a question is asked twice, it deserves a system.
Reproducibility beats cleverness.
Business context before SQL.
Surface uncertainty — don’t hide it.
Don’t automate a bad process.
I’m Anastasia Bogacheva — an analytics engineer for regulated money data: risk, compliance and payment operations at a global payments company, and before that automation and integration in a major bank’s infrastructure department — DWH, ETL and capacity work that was data engineering before I called it that.
I enjoy turning messy analytical questions into systems that answer themselves. I care about reproducible pipelines, honest numbers, and saying “the data can’t tell us that” out loud when it’s true.
Tell me what your data has to survive — auditors, regulators, fraud models — and I’ll reply with how I’d build for it.
kneelisad@gmail.comNOW BUILDING — ANALYST VAULT · AN AI-NATIVE ANALYTICS WORKFLOW ↗