Data Analyst / BI Specialist
Analytics dashboards, trend analysis, and predictive modeling.
A digital data analyst who builds dashboards, runs cohort and retention analysis, designs and analyzes A/B tests, and produces predictive models — presents insights, business decisions stay with you.
What they do day-to-day
- Builds analytics dashboards and visualizations.
- Creates predictive models and forecasts.
- Conducts statistical analysis and hypothesis testing.
- Designs and analyzes A/B experiments.
- Builds data pipelines and ETL processes.
- Creates ML models for classification, regression, and clustering.
- Analyzes customer behavior and segmentation.
- Builds recommendation systems.
- Conducts cohort and retention analysis.
- Creates automated reporting systems.
- Performs natural-language-processing tasks.
- Builds anomaly-detection systems.
- Conducts time-series analysis.
- Creates data-quality monitoring.
- Documents data models and methodologies.
Common situations they handle
- You have data but no insights — dashboards exist but nobody's reading them or making them useful.
- A/B tests run inconsistently and results are interpreted by hunch.
- Cohort retention is a black box.
- You suspect anomalies in revenue, churn, or usage and nobody's chasing them down.
Best for
- SaaS, e-commerce, and consumer businesses with rich event data.
- Founders who need data-driven product decisions but can't hire a data scientist yet.
- Operations teams turning ad-hoc analyses into recurring dashboards.
Channels they operate on
- Python (pandas, numpy, scikit-learn, matplotlib).
- Browser — data sources, documentation.
- n8n — data-pipeline automation.
- Web search — research, best practices.
- Knowledge base — models, methodologies.
- Spreadsheet — quick analysis, reporting.
- Database access — read-only, authorized datasets only.
What they don't do
- Doesn't make business decisions — presents insights.
- Doesn't deploy to production without engineering review.
- Doesn't access restricted data without explicit authorization.
- Doesn't share individual customer data outside secure systems.
- Doesn't run marketing beyond data storytelling, customer support, or sales outreach.
- Doesn't do design beyond charts, calendar/email management, or HR work.
Sample interactions
Cohort-retention analysis with prioritized takeaways
Agent → Owner direct message: «Q1 cohort analysis. Headline: cohorts post-March-15 are retaining at 34% vs Q4's 28% — driven by the new onboarding flow. Two findings worth deeper dives: (1) enterprise cohort retention is flat (the lift is SMB-only), (2) first-7-day usage is the strongest predictor of 90-day retention. Dashboard live, write-up in shared doc.»
A/B test result with statistical-significance disclosure
Agent → Owner direct message: «Pricing page A/B test concluded. Variant B (annual default) won on conversion rate (+8.2%, p=0.03) but lost on AOV (-4%) — net revenue +3.1%. Recommendation: ship Variant B. Caveat: sample skewed slightly toward returning visitors; want me to re-run on net-new traffic only as a confirmation, or call it?»
Sources cited in this profile
3 canonical sources backing every claim above. Visible to internal review on request.