Available — Year 3 CS&E

I build products people actually use.

Engineering · Cloud · Design
Three disciplines. One outcome.

01 — About

One practice.
End to end.

Download CV
UMaT · Computer Science & Engineering · Year 3

I'm Richard Kwaku Opoku. Computer Science & Engineering student. I build software people can actually sign into, pay for, and rely on.

Most of my work sits across cloud architecture, AI, and UI. Currently expanding deeply into cloud security and pentesting. I like seeing the full picture of how systems are secured and put together.

4+Live products
3Disciplines
0Hand-offs
01
Cloud & InfrastructureAWS · GCP · Supabase · CI/CD
02
AI / ML SystemsGemini · OpenAI · RAG
03
Hardware & IoTESP32 · MQTT · Embedded C
04
Product & UI DesignFigma · React · Systems
02 — Stack

What I build with.

Frontend
React Next.js TypeScript Tailwind Vite
Backend
Python Node.js Supabase PostgreSQL Prisma REST APIs
Cloud
AWS GCP Firebase Vercel Docker
AI / ML
Gemini API OpenAI LlamaCloud RAG Langchain
Hardware
ESP32 MQTT Embedded C Raspberry Pi
Design
Figma Canva Framer CSS Three.js
03 — Credentials

Certifications & Hardware.

I

AWS Cloud Practitioner

Amazon Web Services

Cloud ArchitectureSecurityDeployment
Verify Credential ↗
II

Machine Learning Specialization

DeepLearning.AI / Coursera (Andrew Ng)

Machine LearningPythonTensorFlow
Verify Credential ↗
04 — Selected Work

Six products.
Real users.

AI · Scholarships Live ↗

Scholar

Scholarship intelligence platform — from raw documents to personalised funding strategies.

Problem

Students waste hours searching fragmented databases with no eligibility guidance — zero personalisation, zero strategy.

Solution

AI that parses CVs, transcripts & SOPs; matches profiles to real opportunities; tracks deadlines end to end.

Result

Hours of research replaced by a structured, personalised pipeline. Strategy, not just search.

Next.jsGemini APISupabaseLlamaCloudStripe
Schorla app screenshot
AI · Reading Live ↗

InsightFlow

Books turned into actionable intelligence — AI summaries, audio playback, streaks, learning paths.

Problem

People buy books, don't finish them, retain nothing. No tool made reading measurable and genuinely sticky.

Solution

AI-assisted summaries, TTS audio, streak tracking, highlights, personalised recommendations — all in a mobile-first UI.

Result

Reading became a habit with metrics. Users get AI key takeaways, learning paths, and progress analytics.

ReactViteTypeScriptSupabase
InsightFlow app screenshot
SaaS · Data · AI Live ↗

Sheet2SaaS

Your spreadsheet is already your database — we just made it a product.

Problem

Businesses run on spreadsheets but can't share or analyse them without technical overhead.

Solution

Upload any CSV: get charts, filterable tables, AI chat analyst, smart data cleaning, column generation, and public sharing. Free & Pro tiers via Stripe.

Result

Spreadsheets become shareable SaaS dashboards in under a minute with Web Worker performance.

Next.jsSupabaseGemini APIStripe
Sheet2SaaS app screenshot
AI · EdTech Live ↗

StudyMate

AI study assistant — uploads become quizzes, flashcards, summaries, and study plans.

Problem

Students juggle five separate tools — notes, flashcards, quizzes, planner, AI chat — and still feel unprepared.

Solution

One platform: upload materials → AI quizzes, spaced-repetition flashcards, concept glossaries, summaries, planner, and analytics.

Result

Five-tool sprawl collapsed to one. Students get a personalised study loop — upload, learn, test, track.

Next.jsOpenAI APIPostgreSQLTailwind
StudyMate app screenshot
Bot · Campus · AI Web ↗ Telegram ↗

MINEBOTUMaT

Student helpdesk bot for UMaT. Ask it anything about the campus and it answers.

Problem

UMaT students have no quick way to get answers about campus services, schedules, or admin processes without walking to an office.

Solution

A Telegram bot and web interface that handles common student queries, routes them intelligently, and responds instantly.

Result

Students get answers on the spot via Telegram or browser. No queue. No office hours dependency.

Telegram Bot APIPythonReplit
MINEBOTUMaT web interface screenshot
AI · Data Science GitHub ↗

Uni Medi Trend

AI-based analytics dashboard for predicting student clinic visits and illness risk profiles at UMaT.

Problem

Campus health facilities lack actionable, predictive insights into student illness trends, making resource allocation reactive rather than proactive.

Solution

Built a full-stack data pipeline. Trained K-Means models for hostel risk clustering, and Random Forest/Linear Regression models for 4-week visit forecasting.

Result

Deployed a Streamlit dashboard with a MongoDB backend that visuals calendar-year KPIs, diagnosis breakdowns, and predictive illness trends.

StreamlitPythonMongoDBScikit-Learn
Uni Medi Trend Streamlit dashboard
05 — Contact

Have something
to build?

Product idea, cloud architecture, or hardware that needs to talk to the internet.