AI Automation Engineer

Gideon Wafula

Building AI that works while you sleep — voice agents, automations, chatbots, and AI video.

📍 Seoul, South Korea  ·  🇰🇪 From Kenya  ·  Remote-first

The quick version

I'm Gideon Wafula — also known online as Giddy Mufasa. I'm a Kenyan AI automation freelancer currently studying Electrical Engineering at Konkuk University in Seoul, South Korea. Since 2024 I've been building AI systems that help small businesses stop losing leads to missed calls, manual data entry, and slow follow-up.

My clients are typically small business owners — restaurant owners, service-based freelancers, clinics, and agency founders — who know they need AI but don't have a technical team to build it. I come in, scope the problem, build the system, and hand it over working.

What I actually build

AI voice agents: Phone bots built on Retell AI, Vapi, or Bland AI that answer inbound calls 24/7, qualify leads by asking the right questions, and push data to your CRM or send an SMS alert. I pick the platform based on your use case — Retell for fast, low-latency inbound; Vapi for developer teams needing full control; Bland for high-volume outbound at scale.

n8n and Zapier automation workflows: The workflows that eliminate 3–10 hours/week of manual work. Lead capture → CRM → Slack notification → email follow-up sequence, all triggered automatically. I've built these for client onboarding, invoice processing, lead routing, and appointment booking. n8n for teams who want to self-host; Zapier when simplicity matters more than cost.

Website chatbots: Chatbots built on Voiceflow or ManyChat that sit on your website and handle the first conversation — answer FAQs, capture contact info, qualify the lead, and hand off to you when it's worth your time. Restaurants use these for reservations and menu questions. Clinics use them for appointment bookings. Agencies use them for lead qualification.

AI video and faceless YouTube content: Full content pipelines using HeyGen AI avatars — scripted, voiced, captioned, and published on a consistent schedule. Built for business owners who want a YouTube presence without being on camera themselves.

SEO, AEO, and GEO: Getting your business cited by Google, Bing, ChatGPT, Perplexity, and AI Overviews. This includes structured data markup, Answer Engine Optimization (making your content the source AI engines quote), and Generative Engine Optimization (GEO) so your business shows up when people ask AI questions in your niche.

My tech stack

Background

• 2022
  Moved to Seoul, started Electrical Engineering at Konkuk University
  First exposure to serious programming and systems thinking.
• 2024
  Started building AI tools and automation systems
  Learned n8n, Zapier, and the Claude API. Built the first automation workflows for personal use, then started taking on clients.
• 2025
  Launched StyleTry — AI virtual fashion try-on app
  Built a consumer AI product from scratch: styletry-app.vercel.app. Deployed using Vercel, AI image generation APIs, and custom frontend.
• 2026
  Expanded into AI voice agents and AEO/GEO strategy
  Started building on Retell AI and Vapi for small business clients. Published 10+ guides on AI automation, voice agents, and answer engine optimization.

Who I work best with

Small business owners and freelancers who have a specific problem — "I'm missing 40% of my inbound calls", "I spend 6 hours a week manually copying leads from forms into my CRM", "I want to start a YouTube channel but don't want to be on camera." People who want the thing built and working, not a months-long consulting engagement.

My sweet spot is clients in Africa and Asia who want professional AI systems but can't access or afford large Western agencies. I understand the constraints: limited budgets, patchy integrations, and the need for systems that work reliably even when infrastructure isn't perfect.

Outside the work

I'm based in Seoul and genuinely love the city — the food, the efficiency, and the density of interesting people. I care about what AI is going to mean for Africa specifically: not as a market to be sold to, but as a continent that could leapfrog legacy systems the way mobile payments leapfrogged banks. That's the long-term idea behind a lot of what I build.