I am an analytical and self-driven Data Engineer and BSc IT graduate building production-grade data systems from the ground up. My work spans end-to-end pipelines, raw ingestion through transformation, NLP intelligence, ML forecasting, and BI dashboards, with deep focus on Kenya and Africa-wide domain problems in fintech, employment intelligence, real estate, price tracking, public-sector data, and document intelligence.

I have shipped 80+ open-source projects across the full data engineering stack: Airflow orchestration, dbt transformation layers, DuckDB/PostgreSQL storage, HuggingFace NLP models, XGBoost/Prophet forecasting, and React-based data products. Every project is containerised, tested, and built to production standards.

• Batch & Streaming Pipelines: End-to-end workflows using Apache Airflow 3.0 (Task SDK, SimpleAuthManager, asset-based scheduling) for orchestration and Apache Kafka for real-time streaming and decoupled ingestion.
• ELT Architecture & Data Modeling: dbt-driven staging/intermediate/mart layers with automated quality testing (54–72 tests per project) across PostgreSQL, DuckDB, BigQuery, and Snowflake.
• NLP & Intelligence Pipelines: Named entity recognition and sentiment analysis using spaCy, HuggingFace Transformers (FinBERT, zero-shot classification, token classification), and MLflow experiment tracking, applied to news, job postings, legislation, and financial documents.
• ML Forecasting & Predictive Analytics: Time-series forecasting with Prophet (29–109 models per project), regression and classification with XGBoost and SHAP explainability across Kenya, EAC, and Africa-wide datasets.
• Web Scraping & Price Intelligence: Production-grade scrapers using Playwright (session priming for Cloudflare-protected sites) and BeautifulSoup — e-commerce pricing, classified listings, tenders, and property data.
• Cloud & Infrastructure: Containerised solutions on GCP (BigQuery, GCS) and AWS (S3, Lambda, SNS) with Terraform IaC; all local projects run on Docker Compose with isolated environments.
• API Development & Data Serving: Production REST APIs with FastAPI and Pydantic schemas; vector search with pgvector for embedding-based job intelligence.
• Analytics Engineering: Interactive dashboards with Streamlit, Plotly Dash, Apache Superset, Evidence.dev, and Grafana — from fiscal reconciliation to real-time price tracking.

• Portfolio: ian-mwendwa.vercel.app
• LinkedIn: Ian (Dankan) Mwendwa
• Email: idankan571@gmail.com
• Instagram: @faust._