Head of Analytics at Mercuryo Tbilisi, Georgia

Tim Tikhomirov

I build data warehouses and reporting for fintech and crypto companies, set up so your teammates and AI agents can work with the data directly.

I’ve spent seven years building analytics in fintech and crypto, so I know what a well-defined analytics function looks like. I can usually set one up at your company in one to two weeks.

Book a 30-minute call Read a case study

Tim Tikhomirov standing by a mountain lake with autumn larches behind him

Work with me

A data warehouse your team and AI tools can both use

A lot of teams now want to point an AI assistant at their company data. For that to work, the data has to sit in a proper warehouse, and every table and metric needs a description the AI can read. At many early-stage companies neither exists yet.

You might need this if

Your team wants to ask AI about the numbers, but there’s no documented data for it to work from.
Dashboards are slow or keep breaking because they query the production database directly.
You haven’t hired a data person yet, or you just did and they’re stuck building pipelines.

Most of my experience is in fintech, crypto and payments, but good data practice is the same in any industry. The best fit is a company somewhere between seed and Series B.

What’s included

Warehouse A separate database for analytics, so reports and AI queries don’t put load on production.
Pipelines Scheduled syncs from your main sources, with alerts when something fails.
Metrics & dashboards Key metrics defined in one place, and your existing reports rebuilt on top of them.
Semantic layer A description for every table and field. This is what an AI agent reads to understand your data.
AI access Your team can ask an AI assistant questions, and it answers from the warehouse.
Docs in Git Models, tests and documentation under version control, so whoever comes next can pick it up.

Tools I usually work with

ClickHouse
BigQuery
StarRocks
Airflow
dbt
Metabase
Mixpanel
MCP

How a project runs

Call Half an hour. You tell me what data you have and where reporting is falling short.
Proposal A fixed scope and price, in writing.
Build Usually one to two weeks.
Handover I walk your team through everything and leave documentation. Monthly support is optional.

The first step is a 30-minute call. Pricing depends on the scope, and I send a fixed quote once I’ve seen your setup.

Book a 30-minute call

Case study

Dashboards that crashed on a week of data now load in seconds

Moving a payments company’s reporting off the production database

Payments company About 32 hours of work NDA

Executive, finance and operations dashboards ran directly on the production database. Its tables are very wide and hold millions of rows, so any range longer than a week crashed the dashboards.

Reports are more than 10× faster
Reporting no longer touches production
The team can ask an AI agent about the data

Read the full case study

Dashboard load time, before and after
Period	Before	First open	Repeat
24 hours	30+ s	2.5 s	2.3 s
7 days	Crashes	2.3 s	2.0 s
30 days	Crashes	4.6 s	2.5 s
90 days	Crashes	<10 s	1.9 s

Good data should catch something moving while it’s still small, before it’s obvious to everyone else.

The standard I hold my team to, from a Mercuryo interview about spotting trends early and what AI can and can’t do in analytics. Read the interview

Experience

Seven years of analytics in fintech and crypto

Download CV (PDF)

I started in 2018 with web analytics at a digital marketing agency. In 2019 I joined ZION Development in Tbilisi as a data analyst and built the analytics warehouse the company came to rely on for decisions. I then led the analytics team there for four years, mostly on pricing and data architecture.

Since January 2025 I’ve been Head of Analytics at Mercuryo, where I lead the analytics and data engineering teams and set the data strategy.

BSc in Business Informatics with honours, North-West Academy of Public Administration. I work in English and Russian.

2025 – now Head of Analytics Mercuryo
2020 – 2024 Head of Analytics ZION Development
2019 – 2020 Data Analyst / Analytics Engineer ZION Development
2018 – 2019 Web Analyst Digital marketing agency

Selected results

Doubled the number of analytics requests the team resolves each month.
Mercuryo
Cut average BI query time by 72%.
Mercuryo
Moved the data team’s routine work, from metric questions to recurring reports, onto AI tools.
Mercuryo
Added about 20% margin with a dynamic pricing MVP.
ZION Development
Grew B2B purchases up to 3× through competitor-based pricing.
ZION Development
Found insights that led to new products, which eventually grew to over 200,000 MAU.
ZION Development

Contact

Get in touch

The easiest way is to book a 30-minute call. You can also message me on LinkedIn or send an email.

Book a call LinkedIn

Email: