About Me

Data Analyst / Product Analyst

Who I am

Nikita Boyarkin, data / product analyst. Focus — experiments, retention, segmentation. Portfolio of reproducible analytics projects — experimentation, retention, segmentation — with full methodology and code. Built from project work, not from employment.

How I work

From hypothesis to metric. The conclusion can be re-checked: code + SQL + metrics. I prefer reproducible analytics — code, SQL and visualization sit next to every case, so the reader can verify the calculation rather than trust an unsubstantiated number.

My specialization is SQL, Python, visualization, and product analytics with reproducible statistical methodology (CUPED, AA-tests, multiple-comparison correction, ship-gates).

Now

Updated · Sep 18, 2026

Preparing for BI / Data Analyst and Product Analyst mid-level interviews: SQL, pandas, algorithms, mocks.
Maintaining an Obsidian knowledge base on product analytics — a structured Zettelkasten with automation.
Experimenting with AI-assisted analytical workflows on Claude Code + MCP.
Writing about practical data techniques: A/B testing, cohort analysis, RFM segmentation.
Reading: Designing Machine Learning Systems by Chip Huyen.
Open to Data / Product Analyst roles.

How to work with me

A relationship manual — so we don't waste time figuring out the format.

Values

Honesty with data — if the numbers contradict the hypothesis, we change the hypothesis, not the numbers.
Speed of learning — try fast, learn fast, record the result.
Direct feedback — no euphemisms, no "I feel like".

Working style

I like context before a request: why, for whom, what decision the result will drive.
I ask clarifying questions — 5 questions up front beat rework at the end.
I prefer metrics over impressions: "up 6pp" instead of "it got better".

Feedback norms

Direct, specific, with a proposal: what's wrong, why, how to fix it.
Open to criticism — I treat it as free acceleration.
Feedback about the work, not the person.

Meetings

An agenda for any meeting with 3+ people — otherwise it's a chat, not a meeting.
I prefer async: a document + comments over a call for the sake of a call.
Ready for calls with a purpose: a decision, a review, an experiment sync.

What matters to me in a team

Data-driven decisions — metrics as the language, not decoration.
Metric ownership — the team knows what it measures and why.
An experimentation culture — hypotheses get tested, not debated.

Discuss the format → How I can help

Location / format

Moscow. Remote / hybrid — by arrangement. Middle+.

Stack

SQL Python Visualization A/B Testing

Links

GitHub · LinkedIn · CV · Writing · Value · Contact

Message me on Telegram →

GitHub activity

Source: GitHub API · updated 2026-09-19

contributions this year: 694
public repos: 20
followers: 391
current streak (days): 43
longest streak (days): 43
on GitHub since: 2019

Activity over the last 30 days

Daily contributions over the last 30 days — commits, pull requests, issues and reviews in public repositories.

Key takeaways

Over the last 30 days: 414 contributions, averaging 13.8 per day.
Peak activity: 49 contributions in a single day.

Contributions by month

Total contributions per calendar month over the last 12 months.

Key takeaways

Over 12 months: 694 contributions.
The strongest month is Aug with 331 contributions (47.7% of the annual total).

Contributions by week over the year

Total contributions per week (weeks start on Monday) over the last 12 months.

Key takeaways

Active weeks: 26 of 53, averaging 13.1 contributions per week.
The most productive week: 114 contributions.

Contribution types

How the year’s contributions break down by type: commits, pull requests, issues and code reviews.

Key takeaways

686 contributions in total over the year.
The dominant type is Commits (684, 99.7% of contributions).

Top languages

Language share by code bytes across public non-fork repositories. Languages beyond the top 8 are grouped into Other.

Key takeaways

The primary language is Python at 31% of 3109428 bytes of code.
The profile uses 13 languages in total; 9 are shown in the chart.