Data Analyst · Analytics Engineer · Nairobi

I turn messy data into decisions.

Analytics, models and pipelines. Five years in health and pharmacy data, where it has to be right.

~90%

less manual data cleaning, using an AI-assisted workflow

80%

less data fragmentation after merging 3 source systems into one warehouse

500+

pharmacies audited automatically across 4 countries

01 · Work projects

What I shipped for real clients

Anonymized: no client or facility names, no real data in any visual.

Work · healthcare analytics startup

A data platform with no per-seat BI cost

5+ sources and 1.5M records in one warehouse, zero BI licence fees.

Staged pipeline from raw data to analytics-ready tables, with cleaning, patient de-identification and role-based access. Dashboards run on self-hosted Streamlit.

Data EngineeringSnowflakeStreamlit

Work · healthcare analytics startup

AI that turns doctor notes into ICD-11 codes

Manual work down ~90%. Humans review only the ~5% that need it.

An n8n workflow that standardises hospital system data and free-text clinical notes into ICD-11 codes, with a human validation step.

AIData Engineeringn8n

Work · pharmacy tech company

Three systems, one source of truth

Data fragmentation cut by 80%.

Migrated pharmacy, CRM and supply-chain data into one warehouse with 20+ tested dbt models, PR review and a dev-to-prod CI/CD flow.

Data EngineeringdbtSnowflakeRivery

Work · healthcare analytics startup

Stocking a new pharmacy branch with no pharmacy data

A first-month stock list for a branch that had no sales history.

Combined distance to hospitals, gyms and salons, a regional wealth index and other chains' product data, then used KNN to recommend what to stock. Delivered as a client demo.

Machine LearningAnalyticsKNN

Work · pharmacy tech company

Fraud and compliance scoring in 4 countries

Audits automated for 500+ pharmacies.

Designed country-specific composite indexes (normalisation, weighting, scoring), plus a performance analysis of 4,000+ pharmacies.

AnalyticsSQLdbt

Work · pharmacy tech company

Price and volume estimates for 30+ medical products

8 years of trade data, now an automated input for business development.

Modelled historical trade data to estimate price and volume, used to target private-sector opportunities.

AnalyticsSnowflakedbt

02 · Personal projects

What I build on my own time

Public data, open code and live demos anyone can open.

03 · Skills

The toolkit

04 · About

Hi, I'm Mercy.

I started as a data analyst and became an analytics engineer. I lead data at a healthcare analytics startup in Nairobi, where I own the stack from raw data to reporting, and I support a small team. Before that, over three years at a pharmacy tech company, I built the warehouse, models and dashboards that its reporting ran on.

B.Sc. Computer Science, JKUAT. Certificate in Data Science, Moringa School.