Who This Stack Is For

For analysts, product managers, and business owners who want to make data-driven decisions.

What's Included

MCP Servers

PostgreSQL — primary data storage. Transactional database for product metrics.

ClickHouse — columnar database for analytics. Fast aggregations on large datasets.

Grafana — visualization and monitoring. Real-time dashboards, alerts, reports.

SQLite — local analytics. Fast experiments, prototypes, ad-hoc queries.

Skills

Marketing Analytics — analysis of marketing campaigns. ROI, conversions, channel attribution.

Python Developer — scripts for ETL, data processing with pandas, API integration.

Agents

Research Agent — deep data analysis. Insight discovery, anomaly detection, report preparation.

How to Use

1. Connect databases via MCP
2. Describe your business question — Claude will write SQL
3. Create visualization in Grafana
4. Set up automated reports

Example Prompt

Analyze user retention for the last 3 months:
- Cohort analysis by registration weeks
- Breakdown by traffic sources
- Comparison of paid and free users
- Visualization as a heatmap

Analysis Pipeline

Raw data → ETL → Storage → Analysis → Visualization → Insights
    ↓        ↓       ↓         ↓           ↓            ↓
 API logs  Python  ClickHouse SQL      Grafana     Research
 Events    pandas  PostgreSQL Claude   Dashboards    Agent

Results

• Automated data collection and processing
• Interactive dashboards for your team
• SQL queries for any ad-hoc questions
• Regular reports with insights