Skip to content
balisa_
About
Abdoulie Balisa

Statistics student. Systems builder. I turn statistical theory into working tools.

I am finishing a BSc in Statistics at KNUST, and I am a Mastercard Foundation Scholar. I build things, usually because I ran into a question I could not answer with what already existed.

The Gambia's population figures come from the UN and were set before the 2024 census, so I tried an independent Bayesian projection to see what a different method would give. Banks in West Africa cannot share customer data, and rural borrowers were never collected in the first place, so I built a synthetic-data engine that does not need a source dataset. Qur'an apps treat 6,236 verses as a flat list, and the corpus cross-references itself, so I built a galaxy that reorganises around a question.

I work in Python and TypeScript. I am reading Lee-Carter, Cox, and the Basel II framework alongside my coursework, toward actuarial exams. Right now I am applying for internships in data analytics, statistics, or actuarial science.

See the days I shipped
Get in touchSee the workRequest the CV

Stack

AI engineering

What I actually build with

  • LLM orchestration
  • RAG / retrieval
  • pgvector
  • sentence-transformers
  • Prompt specification
  • Tool calling
  • Model fallback chains
  • Output validation
  • NVIDIA NIM
  • Groq

Machine learning & statistics

Modelling, mostly Python

  • PyTorch
  • CTGAN
  • scikit-learn
  • PyMC / MCMC
  • Lee-Carter
  • lifelines
  • HDBSCAN
  • UMAP
  • Pandas
  • NumPy
  • R
  • SQL

Systems & delivery

Getting it in front of people

  • Python
  • TypeScript
  • Next.js
  • FastAPI
  • PostgreSQL
  • Prisma
  • Supabase
  • Chrome MV3
  • Vercel
  • Docker
Experience

Experience

  1. Statistics student, systems builder

    · Independent

    2024 - present · Remote

    • Building production AI products end to end: ingestion, modelling, orchestration, and the UIs on top.
    • Shipped VANTAGE, Gambia Legal Aid, Dalasi Pulse, FORGE, ColdPilot, and AYAT. Six production products built end to end, spanning AI agents, forecasting, and RAG systems.
    • Open-sourcing tooling around RAG, agentic pipelines, and forecasting on the Dalasi.

  2. Freelance Web Developer

    · BS Real Estate

    Jun 2026 · contract · The Gambia

    • Built a real estate website with a private admin dashboard for a property firm in The Gambia, so the team manages their own listings without touching code.
    • Next.js 16 and Prisma 7, a blue and gold brand, admin-only login, and real data throughout, no placeholder content.
    • Delivered the whole thing end to end: the design, the listings dashboard, and the deployment.

  3. Peer Tutor

    · TechUp Africa

    2024 · Remote

    • Mentored learners through the data science curriculum: Python, pandas, statistics, and project reviews.
    • Ran code-review sessions and helped unblock learners on ML fundamentals.

  4. AI Automation Developer

    · Lujo Heights Real Estate

    Mar 2024 · 3-week contract · Remote · Nigeria

    • Short-term engagement to automate their manual lead pipeline, built a Python script that pulled CRM data and tagged prospects as Hot, Warm, or Cold based on engagement signals.
    • Wired up n8n workflows to route leads automatically so the sales team stopped doing it by hand.
    • Left them with a working system and a brief doc, done in three weeks, no ongoing dependency.
Education

Education

2022 - present

BSc Statistics

Kwame Nkrumah University of Science and Technology (KNUST)

Kumasi, Ghana

Probability TheoryStatistical InferenceRegression AnalysisLinear AlgebraStochastic ProcessesR programming
Certifications

Credentials

9 certificates spanning AI engineering, data science, and software engineering. Tap any to verify.