What this is
Inside a typical company, data lives in 5-10 places: CRM, 1C, ad platforms, Google Analytics, Excel reports from employees. To answer a simple "which ad channel actually pays back," an analyst spends two days joining tables. By the end of the week the data is already stale.
We build an ETL pipeline that once a day (or in real time, if needed) pulls everything into a single warehouse. On top we put Metabase or custom dashboards - and the same questions get answered in 2 minutes instead of 2 days.
A separate strand of work is product analytics for SaaS: cohorts, retention, funnels. We help you see not just "revenue dropped," but "why users from the March cohort stopped coming back."
Stack
What's included
- ETL pipeline from your systems
- Unified data warehouse
- Dashboards in Metabase / custom
- Cohort analysis and retention
- A/B test infrastructure
- Alerts on key metrics
Typical tasks
The kind of requests that usually come in for this service.
- Management dashboard for the CEO
- Sales funnels and retention
- SaaS product analytics
How we work on this service
Four stages with clear deliverables
This breakdown is specific to this service - the general methodology lives on the "How we work" page.
Source audit
We map every system that holds data and look at volumes and quality.
Deliverable
Source registry + ETL plan
Data warehouse
We build a single store: Postgres for small volumes, ClickHouse for large.
Deliverable
Warehouse with historical backfill
Dashboards
Metabase or custom dashboards tailored to your key metrics.
Deliverable
3-5 dashboards for different roles
Alerts
We wire up Slack/Telegram notifications when a metric goes out of range.
Deliverable
Alerts on critical KPIs
When this isn't your task
We don't take these cases under this service label - honestly, so you don't lose time.
- Projects with less than ~10k rows of data - Excel + pivot tables usually cover that
- Scoring and ML models - that's a separate task, not BI; we take it only if you have no in-house DS team
- Analytics on someone else's data (competitor scraping, etc.)
FAQ
Common questions about this service
Metabase vs Power BI?
Metabase is faster and simpler for technical teams, open-source, can be hosted on-premise. Power BI is visually richer but tied to the Microsoft ecosystem. We use Metabase more often - but if you're already on MS, we can work with Power BI too.
Do we need a dedicated BI engineer?
During build - no, we do everything. After launch - depends on complexity. Often an analyst who can write SQL and work with Metabase is enough.
What about personal data?
We keep PII separate from the analytical warehouse, with anonymisation / hashing at the ETL stage. Compliance with Kazakhstan's personal-data laws is handled at the infrastructure level.
Other services