👔 Backend & Linux Systems Engineer at Sendhome Financial Services | Ex Software Developer at Africom Holdings
🏫 I earned a Bachelors Of Technology Honours Degree In Software Engineering from the Harare Institute of Technology
👨🏾💻 Currently working with: Java, Spring Boot, Python, Vue and Linux Systems
📚 Personally using and upskilling in :-
- AI:
LLM Agents,Conversational AI,Computer Vision,Ollama - Signalling/Networking:
RestApi,GRPC,ISO 20022,Kafka - Infrastructure:
Docker,Ansible,Azure,Monitoring (Grafana Stack, Zabbix) - Architecture:
TOGAF
| Project | What it is | Stack |
|---|---|---|
| ERPNext PWAs | Four mobile-first Sales, HR, Accounts and Admin apps on top of ERPNext and Frappe HR v16 for Southern African small businesses (PAYE payroll, USD/ZWG, bank reconciliation) | Vue 3 Frappe Tailwind Vite |
| FortKnox | Event-driven ISO 20022 payments platform: JWT gateway (ZW), pacs/acmt processor (Capital), reactive double-entry wallet ledger (VaultView), all deployed by an Ansible orchestrator | Spring Boot Kafka WebFlux Ansible |
| Logikbrain | Natural-language analytics agent for ERPNext: plain-English questions become live queries, charts and one-line insights, with a self-correcting retry loop for hallucinated columns | Python LLM Agents Ollama Claude |
| WhatsApp Customer-Care Agent | Guardrails-first LLM support agent for a mobile network: rule-based routing of high-risk requests, grounded bundle answers, PII redaction, human handoff and golden-set evals | FastAPI Mistral NeMo Postgres Presidio |
| Plant Disease Detection | Published BTech capstone: SSD MobileNet leaf localisation + ResNet50 disease classification at the edge on an NVIDIA Jetson Nano | PyTorch CUDA Flask |
- Published: Zawala, K. & Chibaya, Y. Disease Detection System: A Technology-Based Approach to Plant Disease Management. Software Engineering Department, HIT
- Entelect Hackathon: finished 85th of 489 (peaked at 56th) on a multi-species garden optimisation problem, solved in
Python


