Francis Kofigah

An AI-enhanced digital learning platform designed to make learning engaging, measurable and collaborative.

Students learn through interactive content, complete assessments, compare progress against peers through benchmark rankings, and receive AI-assisted learning support tailored to their academic journey.

• Product architecture
• Full-stack development
• User experience design
• AI feature integration
• Learning workflow design

Mobile School Report Card — a national education platform that has evolved continuously since 2016.

In 2026 I led the complete redesign of the mobile experience using AI throughout the design and engineering process — rethinking years of accumulated workflows while preserving operational continuity for existing users.

AI became part of the engineering process itself: analysing user journeys, challenging interface decisions, generating implementation strategies, reviewing edge cases and accelerating development.

An AI legal research assistant that helps legal professionals retrieve grounded answers from trusted legal sources instead of relying solely on language models.

Building it required learning Retrieval-Augmented Generation, vector databases, embeddings and Python to build a system that retrieves relevant legal material before generating responses.

An AI-powered platform supporting African nurses through the NCLEX migration journey.

Combines intelligent guidance, personalised checklists, reminders, payment workflows and communication automation into a single experience that reduces the complexity of international migration.

Turning a stock exchange nobody watches into a daily signal.

An automated pipeline that scrapes Ghana Stock Exchange data, runs it through an LLM to generate plain-language market summaries, and delivers a daily digest — for free, at zero server cost.

Internal software supporting customer engagement, online ordering, marketing automation and operational workflows for a growing consumer brand.

Running the business provided an opportunity to build software while continuously learning from real customer behaviour and operational challenges.

Ten years of WhatsApp chats, turned into structured data.

A two-stage parsing pipeline — regex first, LLM second — that converts a decade of raw WhatsApp order history into clean, structured records, now also powering customer-facing delivery documents.