Skip to content
DM_

Data Engineer & Data Scientist

DavidMwandairo.

I design pipelines, models, and dashboards that survive contact with real data, from raw API JSON to boardroom-ready answers.

View case studiesRésumé

Open to data engineering & science roles

  • Python
  • SQL
  • PostgreSQL
  • ClickHouse
  • dbt
  • Airflow
  • Kafka
  • Debezium
  • DuckDB
  • Docker
  • Terraform
  • FastAPI
  • scikit-learn
  • Power BI
  • Grafana
  • Great Expectations
  • Prometheus
  • Git

Featured work

Systems that shipped.

Four builds, told honestly: what broke, what I decided, and what it produced.

MarketFeed Lakehouse: A production-grade medallion lakehouse that turns rate-limited market APIs into an auditable, tested analytics warehouse.
PythonAirflowdbtDuckDB

MarketFeed Lakehouse

Medallion architecture over Alpha Vantage and FRED data: immutable bronze, SCD Type 2 dimensions, and point-in-time-correct marts, orchestrated by Airflow and modeled in dbt.

Read case study
DebeziumKafkaClickHousedbt

Fineract Analytics Platform

Apache Fineract → PostgreSQL → Debezium/Kafka → ClickHouse → dbt marts and ML features, orchestrated by Airflow with Prometheus/Grafana observability. The whole stack starts with one command.

Read case study
SignalTel Churn Intelligence: An end-to-end churn model for a telecom: leakage-safe pipelines, honest model selection, and a production FastAPI service.
scikit-learnFastAPIMachine LearningDocker

SignalTel Churn Intelligence

7,021 customers scored; logistic regression chosen over gradient boosting for recall; F1 0.6129 and ROC-AUC 0.8397, served through a documented FastAPI service in Docker.

Read case study
Tembo Hotel Analysis: From a dirty hotel spreadsheet to a Power BI board the Director can trust, with every headline tied to a SQL query.
SQLPostgreSQLPower BIDAX

Tembo Hotel Analysis

286 messy bookings cleaned to 285 in PostgreSQL, six business questions answered in SQL, and a three-page Power BI dashboard with 25 DAX measures.

Read case study

0

flagship builds

0

customers scored

0.00

ROC-AUC, churn task

0

DAX measures shipped

What I do

Four disciplines, one pipeline.

01

Data Engineering

Batch and streaming pipelines that treat raw data as an audit trail: Airflow-orchestrated ingestion, CDC with Debezium and Kafka, idempotent loads, and quality gates before anything reaches a consumer.

AirflowKafkaDebeziumDockerTerraform
02

Analytics Engineering

Warehouse modeling with dbt: staging → intermediate → marts, slowly-changing dimensions done properly, point-in-time-correct joins, and tests that make the grain explicit.

dbtPostgreSQLClickHouseDuckDBGreat Expectations
03

Machine Learning & Data Science

Leakage-safe scikit-learn pipelines, honest model selection (F1 over accuracy on imbalanced data), and serving through documented FastAPI services with tests and containers.

scikit-learnFastAPIpytestDocker
04

Dashboards & Storytelling

SQL that answers the actual business question, then Power BI and Grafana surfaces where every headline number ties back to a query someone can read.

SQLPower BIDAXGrafanaPrometheus

The full story is in the work: read the case studies or email me about yours.

Next step

Have data that misbehaves?

I'm open to data engineering and science roles. Email reaches me fastest.

Email meAll contact links