I build reliable, production-grade software that solves real-world operational bottlenecks for growing businesses, startups, and NGOs. My focus is on automated workflows, secure infrastructure, and practical AI that saves time and reduces manual errors.
Currently: Founder & Software Developer at Ithoka Microsystems
I specialize in end-to-end full-stack development — taking complex requirements and turning them into accessible, secure tools.
- Practical AI Integrations: Building automated data extraction and processing pipelines (RAG, LLMs, Computer Vision) that actually work in production environments.
- Secure Web3 Infrastructure: Developing smart contracts, wallet integrations, and privacy-preserving payment rails (EVM, Solana).
- Workflow Automation: Replacing manual back-office tasks with secure, reliable backend microservices.
- Full-Stack Delivery: Crafting clean, responsive user interfaces (React/Next.js, Typescript) backed by robust backends (NodeJS, Python, Go, Rust).
| Project | Description | Stack |
|---|---|---|
| AutoBooks Finance | Automated bookkeeping engine that combines AI receipt parsing with secure Web3 payment rails for SMEs. | Python (Django/FastAPI), Next.js, Node.js Web3 |
| Agrosight AI | Edge-deployed crop diagnostic tool utilizing computer vision, designed for low-bandwidth rural networks and WhatsApp API. | Python, YOLOv8, Next.js |
| Camouflage Protocol | Privacy-preserving B2B payment infrastructure on Solana that protects sensitive corporate transaction data. | TypeScript, Solana, Cloak ZKP |
| Talos (OpenClaw Vault) | Secure execution layer allowing AI tools to safely automate internal B2B workflows without exposing root API credentials. | Go, Auth0, Next.js |
| Hungerless Engine | Predictive logistics and geospatial routing engine built for anticipatory humanitarian aid deployment. | Python, Geospatial Mapping |
Reliability over novelty Tools should solve problems, not create them. I build systems designed to handle edge cases gracefully and operate predictably under all conditions.
Secure by default Whether handling business data or digital assets, security, key management, and access control are built-in from day one.
AI as a practical utility, not magic LLMs and generative models are powerful, but they require strict deterministic guardrails and validation to be trusted in live business operations.
Backends do the heavy lifting Complex validation, transaction signing, and security logic belong on the server, keeping client applications fast, secure, and lightweight.
- Workflow automation and data processing for SMEs and NGOs
- Practical Web3 integrations (Payments, Smart Wallets)
- Full-stack React (Next.js) + Node/Python engineering
- Computer vision and edge-deployed AI solutions