Patrick Ndieka
Selected work

Projects, framed by Stack → Architecture → Deployment.

Every build below breaks down the same way: what it's built with, how it's structured, and how it ships.

Shipped
3
Cloud
AWS
Core Stack
Py / TS
Filter by stack:
Layer 01 — The "Live Model" project

Kenyan Market Trends Dashboard

A Streamlit dashboard that processes a real-world dataset — Kenyan market trends — hosted end to end on AWS.

Stack
PythonStreamlitpandasAmazon EC2
Architecture

A nightly batch job pulls source data into S3; Streamlit reads the processed parquet files and renders interactive views.

Deployment

Containerized with Docker, deployed to an EC2 instance behind Nginx, monitored with CloudWatch.

Layer 02 — The "SaaS" project

ML-Powered Insights Platform

A Next.js + FastAPI application that runs a machine learning model to surface insights for its users.

Case study coming soon GitHub
Stack
Next.jsFastAPIPostgreSQLscikit-learn
Architecture

FastAPI serves model predictions behind a REST API; the Next.js App Router renders results with server components.

Deployment

API shipped as a container image on AWS Lambda; frontend deployed separately with CI on every push to main.

Layer 03 — The "Infrastructure" project

Secure Data Environment (IaC)

A GitHub repository of AWS CloudFormation and Terraform scripts that stand up a secure, scalable data environment.

Stack
TerraformAWS CloudFormationAmazon VPCAWS IAM
Architecture

Modular Terraform stacks provision a VPC, private subnets and IAM roles scoped to least privilege for data workloads.

Deployment

Applied through GitHub Actions with a plan → approve → apply pipeline; state stored remotely in an encrypted S3 backend.

Have a high-stakes data or cloud challenge?

Initiate
connection.

Projects, contract work or a quick technical question — reach out directly.

[City, Country]