Accra, Ghana

Samuel
Ankapong.

AI Strategist & Tech ConsultantSoftware Developer

I help businesses decide where AI is useful, build practical solutions, and check whether they improve the work.

View my résumé PDF

From a business problem
to a working solution.

I’m a software developer offering AI strategy and technology consulting. I help small businesses and teams make sense of a workflow, choose a practical approach, and build the software to put it to work.

At Future GH, I lead systems and development for African SMEs and entrepreneurs. I work across the frontend, backend, and the decisions that connect them.

I’m also a Biomedical Engineering graduate from the University of Ghana. Alongside product development, I work on machine learning, biomedical signals, and model evaluation. You can explore that work in my ML & research portfolio.

Start with the task.
Not the tool.

For small businesses, agencies, and teams with repetitive work or a product to build. Start with a focused review, or bring a defined software brief.

AI strategy & workflow review

Review one process, compare AI with simpler automation or existing software, and decide what is worth building.

You get: a workflow map, practical options, risks, and a scoped pilot plan.

AI integration & automation

Build a bounded assistant, research workflow, chatbot, or integration around a specific task—with people in control of important actions.

You get: a working pilot, agreed tests, documented limitations, and handover.

Software development & technical consulting

Turn requirements into business websites, web and mobile applications, dashboards, and connected systems. AI is an option, not a requirement.

You get: an agreed implementation scope, working software, and documentation.

Scope comes first. Ongoing support and improvements can be agreed separately after delivery.

Tell me about the task

The work behind the offer.

01 — 06

  1. 01 Research brief
  2. 02 Public sources
  3. 03 Structured output
  4. 04 Human review

Process overview · no client records shown

02 / Research & operationsReusable automation

CDJ — lead research automation

A reusable browser-assisted research tool I built with Chrome DevTools and the Jev API. I’ve used it for a client’s school-lead research and my own business prospecting.

  • Browser automation
  • Lead research
  • Reusable workflow
Contribution & boundaries

The task: research prospects against a brief rather than collect an unfiltered list of contacts.

My contribution: built the reusable skill connecting browser tooling with the Jev API, then applied it to client and personal research.

Evaluation focus: relevance, source quality, duplicates, and manual validation. A research record is not proof of buying intent; conversion or time-saving results are not claimed.

Privacy: client lead sheets are not published here. The diagram above is a process overview, not a client dataset.

Skulbii's school management website showing its offline-first school records productVisit project
03 / EducationWeb application

Skulbii

School records that don’t depend on a connection. An offline-first application for attendance, assessments, and fees, with automatic syncing when back online.

  • Offline-first
  • Role-based access
  • Audit logs
Contribution & technical choice

The task: keep school-record entry usable when connectivity is unreliable.

My contribution: built the school-records application with offline entry, syncing, role-based access, and audit logs.

The choice: focus on dependable records and access rather than adding AI to every workflow. The product also covers term reports and admissions for schools in Ghana from KG to SHS.

Fashley's fashion design website with fabric-inspired visual conceptsVisit project
04 / Fashion & designAI application

Fashley

See what your fabric could become. A design application that turns fabric photos and style preferences into outfit, bag, and shoe previews.

  • Image generation
  • Saved designs
  • Tailor briefs
Behind the build

I built the application for generating fashion concepts from uploaded fabric photos. Users can save designs, download previews, and prepare a brief for a tailor.

The output is a visual concept, not a sewing pattern. A maker still needs to confirm fabric suitability, construction, and fit.

SWIM Talent Hub's website connecting students with paid company projectsVisit project
05 / Work & opportunityTalent platform

SWIM Talent Hub

Connecting students with paid company projects in freelance work, product testing, and system evaluation.

  • AI matching
  • Student & company dashboards
  • Chatbot
Behind the build

I built and maintained the platform’s frontend and backend, including dashboards for students and companies. I also developed an AI matching engine to connect students with relevant opportunities and integrated a chatbot.

06 / EducationNext.js

PEVAL

A responsive peer-evaluation application, built with Next.js so students can evaluate their peers through a web interface.

  • Next.js
  • Node.js
  • Express.js

A useful result.
Not just a convincing demo.

  1. Understand the workflow

    Look at the task, people, systems, and constraints. Agree on the problem and what success would mean.

  2. Choose the right approach

    Compare AI, simpler automation, and existing software. Sometimes the right recommendation is not to automate.

  3. Build and test

    Implement a bounded solution. Check outputs, exceptions, human review effort, and running costs against agreed criteria.

  4. Hand over and improve

    Document how it works, its limitations, ownership, and maintenance. Decide together whether to improve, expand, or stop.

Where I’ve contributed.

Feb 2025 — Present

Systems & Development Lead

Future GH

Leading development of web applications and business systems for African SMEs and entrepreneurs. Defining application architecture and overseeing implementation across websites, automation, and AI workflows.

Oct 2024 — Aug 2026

Full Stack Developer

SWIM (Student Workforce Innovation Movement)

Built and maintained the frontend and backend, including student and company dashboards. Developed an AI matching engine and integrated a chatbot.

Sept 2024 — Feb 2025

Web Developer

African Intelligence

Designed and built landing pages and digital assets for marketing campaigns, working with product teams on brand and user-experience requirements.

Feb 2024 — Apr 2024

IoT Intern

Techten Planet GH

Built a React Native mobile app for controlling a Raspberry Pi 4 smart home system. Tested app commands against device responses to verify control of connected appliances.

View full résumé
  • JavaScript
  • Python
  • React
  • React Native
  • Next.js
  • Node.js
  • FastAPI
  • SQL
  • Git
  • Docker
  • AI workflow automation