...///

SamuelWanyua

AI Engineer · Software Engineer · Founder, Sauticare

|
View Work →My Resume ↓
SCROLL
Samuel Wanyua

Breaking barriers with
Accessible innovation.

I am an AI Engineer & Software Developer building accessible, human-centered technology. I specialize in speech recognition, NLP, and machine learning — with a focus on making AI work for underserved communities. I bridge the gap between research and real-world deployment — from fine-tuning models to shipping production-ready products.

Currently building Chacha Speech — a speech therapy companion for children with communication disabilities, and developing bilingual ASR systems for Kenyan English and Swahili. Winner, HackAbility 2026.

HackAbility 2026 Winner — AI for Accessibility
Runner's up ASR Innovation Sprint
Co-Founder Chacha Speech
Founder — Sauticare

Technical Trajectory

Sauticare

Founder

Nairobi, Kenya

July 2025 – PresentStartup
  • Tuned the ASR pipeline to respond in under 500ms, with solid support for Kenyan English.
  • Built a pronunciation scoring system using WER and CER so learners get real, measurable feedback on their progress.
  • Wired up a FastAPI + Supabase backend handling lesson delivery, progress tracking, and automated milestone rewards.
  • Shipped a Next.js 15 frontend with gamified achievements that pushed simulated learner engagement up by 25%.

NW Realite

Data Analytics Engineer

Nairobi, Kenya

Nov 2025 – Feb 2026Contract
  • Designed OCR and RAG-based pipelines to extract structured data from unstructured property valuation reports.
  • Built scalable ETL workflows using Python and loaded transformed data into Supabase databases, improving performance.
  • Developed Power BI dashboards supporting predictive and prescriptive analytics for executive-level strategy.
  • Translated complex analytics outputs into executive-level insights, bridging the technical-business gap.

Freelance

Software Engineer & AI Engineer

Remote

Jan 2025 – PresentFreelance
  • Designed end-to-end bilingual ASR speech tutor for Kenyan English and Swahili using Whisper and Wav2Vec2.
  • Optimized WER/CER metrics and latency for offline AI deployment in low-resource rural Kenyan classrooms.
  • Built HemaHaus: a price prediction system (91.9% accuracy) across 190K+ listings using Scikit-learn and FastAPI.
  • Engineered early-stage RAG pipelines using embeddings and vector databases for unstructured property insights.
  • Managed full ML project lifecycle from problem scoping to deployment across multiple concurrent clients.

Innovex Solution Limited

Software Developer

Nairobi, Kenya

Aug 2024 – Jan 2025Contract
  • Built insurance management software using Next.js, React, TypeScript, Redux, and Tailwind CSS.
  • Improved client operational efficiency by 10% through frontend architecture optimization.
  • Refactored legacy codebases and resolved complex UI/UX issues to improve cross-device responsiveness.

CodePerfect Solutions

Technical Writer — Data & AI

Remote

May 2025 – PresentPart-time
  • Write in-depth technical articles on Data Engineering and AI (ETL, Model Deployment, MLOps).

Projects I've worked on

Founder · Live
Real-Time ASR Feedback Gamified Learning

Sauticare

Inclusive bilingual (Kenyan English + Swahili) speech recognition tutor with gamified learning and adaptive pronunciation scoring. Designed for learners with speech impairments. Winner of HackAbility 2026.

FastAPISupabaseNext.js 15WhisperWav2Vec2
Explore →
Co-Founder · Software Developer
Interactive Speech GamesPractice & Feedback

Chacha Speech

AI-powered speech therapy companion for children with mild to moderate speech impairments (stuttering, lisping, apraxia). Bridges the 'practice gap' with real-time analysis and gamified exercises. Revenue through school/clinic subscriptions.

ASRTTSFastAPINext.jsSupabase
Github Repo →
AI/Software Engineer (CDLI Makerthon)
100% OfflineEdge Deployed

VoiceNote

Offline-first, edge-deployed speech therapy system on Raspberry Pi 5 for low-resource environments. Delivers structured sessions with TTS prompts and real-time pronunciation scoring. No internet required. Built during HackAbility 2026.

WhisperWav2Vec2Raspberry Pi 5Edge MLTTSPython
Learn More →

Technical Expertise

Languages

PythonTypeScriptJavaScriptSQL

Frontend

ReactNext.jsReduxTailwind CSSMaterial UIStreamlit

Backend & APIs

FastAPINode.jsREST APIs

Databases

PostgreSQLSupabaseMySQLMongoDBBigQuery

Machine Learning & AI

PyTorchTensorFlowScikit-learnXGBoostTransformersWhisperWav2Vec2NLPDeep LearningLLMsASRAI AgentsAI Workflows

MLOps & Deployment

DockerCI/CDGitHub ActionsHugging Face SpacesGCPVercelRailwayHerokuRender

Data & Analytics

ETL PipelinesOCRData ModelingFeature EngineeringFine-tuningWER/CERPower BILooker StudioPlotlyMatplotlibSeaborn

UI/UX & Design

FigmaAdobe IllustratorAdobe PhotoshopAdobe InDesignCanva

Tools & Workflow

GitGitHubn8nPostmanVS CodeJupyter NotebookNotionClickUpAsanaTrello

Articles on Medium

The Machine Learning Landscape
M
Machine Learning

The Machine Learning Landscape

Read on Medium →
Building Trust Through Responsible AI
M
Responsible AI

Building Trust Through Responsible AI

Read on Medium →
ACID in Databases
M
Databases

ACID in Databases

Read on Medium →
OLTP vs OLAP
M
Data Systems

OLTP vs OLAP

Read on Medium →
Introduction to Data Pipelines
M
Data Engineering

Introduction to Data Pipelines

Read on Medium →
Choosing Technologies Across the Data Engineering Lifecycle
M
Data Engineering

Choosing Technologies Across the Data Engineering Lifecycle

Read on Medium →
A Modern Data Infrastructure
M
Infrastructure

A Modern Data Infrastructure

Read on Medium →
Common Data Pipeline Patterns: ETL, ELT & ETLT
M
Data Engineering

Common Data Pipeline Patterns: ETL, ELT & ETLT

Read on Medium →
Designing Good Data Architecture
M
Architecture

Designing Good Data Architecture

Read on Medium →
The Data Engineering Lifecycle
M
Data Engineering

The Data Engineering Lifecycle

Read on Medium →
The Rise of AI Engineering
M
AI Engineering

The Rise of AI Engineering

Read on Medium →
The Machine Learning Landscape
M
Machine Learning

The Machine Learning Landscape

Read on Medium →
Building Trust Through Responsible AI
M
Responsible AI

Building Trust Through Responsible AI

Read on Medium →
ACID in Databases
M
Databases

ACID in Databases

Read on Medium →
OLTP vs OLAP
M
Data Systems

OLTP vs OLAP

Read on Medium →
Introduction to Data Pipelines
M
Data Engineering

Introduction to Data Pipelines

Read on Medium →
Choosing Technologies Across the Data Engineering Lifecycle
M
Data Engineering

Choosing Technologies Across the Data Engineering Lifecycle

Read on Medium →
A Modern Data Infrastructure
M
Infrastructure

A Modern Data Infrastructure

Read on Medium →
Common Data Pipeline Patterns: ETL, ELT & ETLT
M
Data Engineering

Common Data Pipeline Patterns: ETL, ELT & ETLT

Read on Medium →
Designing Good Data Architecture
M
Architecture

Designing Good Data Architecture

Read on Medium →
The Data Engineering Lifecycle
M
Data Engineering

The Data Engineering Lifecycle

Read on Medium →
The Rise of AI Engineering
M
AI Engineering

The Rise of AI Engineering

Read on Medium →

Let's Build Something

Open to AI/ML projects, accessibility-focused collaborations, speaking opportunities, and mission-driven roles.

GitHubGitHubLinkedInLinkedInEmailEmail

“I reply fast. Probably faster than my ASR inference latency.”