Building voice AI that speaks African languages.
I'm Takura Mukaro — a software engineer and entrepreneur in Harare, Zimbabwe. Co-founder & technical lead at Nerva Studios, where I do the deep work: TTS quality, Shona phoneme injection, and building for the Bantu languages the big models forget.
Selected work
2023 — 2026 / 05 PROJECTS
Teaching machines to hear and speak the languages they were never trained on.
VoiceQL
A voice-to-SQL agentic pipeline — talk to your database in natural language. Grew from a production prototype built with Pipecat, Whisper, GPT-4o, LangChain and SQLite. Supervised by Dr. Mkhuseli Ngxande, co-supervised by Dr. Ndivhuwo Makondo (IBM Research Africa).
Shona speech synthesis
Optimizing TTS output quality for a language the major models don't support natively — Kokoro TTS with Shona phoneme injection through IPA syntax, working around the absence of native Shona/Bantu coverage.
Toolkit
FULL-STACK / VOICE-AI / MLPython · FastAPI · NestJS · Node.js · Java · C/C++ · microservices · OAuth2/JWT/RBAC
React · Next.js · TypeScript · Tailwind · React Native
Stripe · Flutterwave · PayNow · Reloadly · PayPal · multi-currency wallets · settlement pipelines
ASR/TTS · Whisper · Deepgram · ElevenLabs · Kokoro · LangChain · Anthropic SDK · PyTorch
PostgreSQL · Redis · MySQL · Firebase · TypeORM · SQLite
Docker · GitHub Actions · Jenkins · DigitalOcean · GCP · Azure
Evidence-based solutions to real industry gaps.
I'm a software engineer with a passion for building things that close real gaps. Over 4+ years I've moved from applying existing knowledge to developing my own solutions — most recently in voice AI for underserved languages.
Native Shona speaker. That's not incidental to my work — it's central to it. From neural style transfer in my undergrad thesis to a voice-to-SQL agent for my Master's, the thread is the same: make powerful systems work for people they usually leave out.
Education
Certifications
Awards
Languages
Earlier
Let's build something.
Open to freelance work and research collaborations in integration engineering, voice AI, API development, and system optimization. If you've got something worth building, let's talk.