Full-stack engineer in Accra, Ghana. Python and FastAPI on the backend, React and Next.js on the front, PostgreSQL underneath.

Most of what I have built is not public. Here is what it was.

Sole engineer on a financial data and wealth management platform for the Ghanaian capital markets, from first commit to production. Automated ingestion from four public sources with gap detection and backfill, a live-price portfolio valuation engine with daily P&L and AUM analytics, and a RAG investment assistant on the Anthropic Claude API with pgvector.

160 backend modules, 138 API routes, 21 data models, 31 frontend pages, 35 automated test files.

Four internal systems replacing paper record keeping: a digital flight logbook, personnel and training records, officer course tracking, and a training directorate management system. Closed network, so there is no code and there are no screenshots. The flight logbook is still in use, and I was still shipping features to it six months after my national service ended.

• stockalert - pharmacy inventory with batch-level expiry tracking, FEFO dispensing across batches in a single transaction, and barcode scanning. React 19, Express, Prisma, Vitest.
• Dermaid - nine-class skin disease classifier using MobileNetV2 transfer learning. Final year project, KNUST.
• GAN-FOR-SKIN-COLOUR
  • StarGAN in PyTorch, synthesising skin-tone diverse training data to correct the dataset bias in the classifier above.

Python, FastAPI, SQLAlchemy, Celery, Redis, PostgreSQL, pgvector, Flask TypeScript, React, Next.js, Tailwind CSS Docker, AWS, Cloudflare Workers, GitHub Actions Playwright, Scrapy, BeautifulSoup

Open to remote or Accra-based roles. ronardbotchway@gmail.com