Skip to content
DanielTech
DanielTech

Available for opportunities· Nigeria · Open to remote work

Daniel / Tech

AI Engineer & Software Developer

My work spans applied AI, full-stack software, LLM systems, NLP, computer vision and AI evaluation. I care about more than getting a model or prototype to work once. I like understanding where systems fail, improving them through testing and evaluation, and building software people can actually use.
Applied AI
  • AI-powered applications
  • LLM applications
  • RAG
  • NLP
  • Computer vision
  • AI evaluation
Software
  • Full-stack software
  • APIs & backend systems
  • Automation
Product
  • Product engineering

01Case studies

Flagship projects.

Project 01MeetingBot

MeetingBot

AI application

An AI meeting assistant that turns a transcript or audio recording into a summary, action items, decisions and participants — with a question-answering chat grounded in what was actually said.

  • 79 passing tests
  • OpenAI path with local-model fallback
  • Retrieval failure found and fixed through evaluation
  • Python
  • Flask
  • Whisper
  • FAISS
  • sentence-transformers
  • OpenAI API
  • Hugging Face Transformers
  • +5
Read the case studySource — MeetingBot on GitHub (opens in a new tab)
  1. 01Transcript or audio
  2. 02Whisper
  3. 03Speaker-turn chunking
  4. 04Embeddings + FAISS
  5. 05OpenAI or local QA
  6. 06Meeting intelligence
Fig. 01 — MeetingBot — system overviewDiagram

Project 02Trash2Cash

Trash2Cash

AI product · Recycling platform

An AI-powered recycling platform for Nigeria — estimate a cash reward for recyclables, find where to drop them off, and ask a grounded assistant that understands how Nigerians actually talk about waste.

  • 96.9% grounded-answer accuracy
  • 1.6% raw hallucination rate
  • 64-case live AI evaluation
  • Python
  • Flask
  • OpenAI API
  • NLP
  • pytest
  • JavaScript
Read the case studySource — Trash2Cash on GitHub (opens in a new tab)
  1. 01User message
  2. 02Intent extraction
  3. 03Schema validation
  4. 04Deterministic tools
  5. 05Grounded phrasing
  6. 06Local NLP fallback
Fig. 02 — Trash2Cash — system overviewDiagram

02Approach

How I work.

  1. 01

    Measure it, don’t assume it

    Tests and evaluation sets come with the feature, so failures are found deliberately rather than by users.

    See: MeetingBot retrieval fix
  2. 02

    Keep facts out of the model’s hands

    Where correctness matters, deterministic code computes the answer and the language model only explains it.

    See: Trash2Cash grounding
  3. 03

    State the limitations

    Every case study lists what the system can’t do yet. Knowing the edges is part of the engineering.

    See: MeetingBot limitations

03Selected repositories

Smaller builds and experiments.

All repositories on GitHub (opens in a new tab)

04Stack

Tools and technologies.

01Applied AI
  • AI engineering
  • LLM systems
  • RAG
  • NLP
  • Computer vision
  • AI evaluation
02Software Engineering
  • Full-stack development
  • Backend & API development
  • REST APIs
  • SQL
  • PostgreSQL
03Languages & Frameworks
  • Python
  • TypeScript
  • JavaScript
  • Flask
  • FastAPI
  • Node.js
  • Next.js
  • React
  • Tailwind CSS
04ML Frameworks
  • PyTorch
  • TensorFlow
05Tools & Infrastructure
  • Git
  • GitHub
  • Docker
06Also
  • Data analysis
  • Technical writing

Available for opportunitiesNigeria · Open to remote work

Need an AI product built properly?

omotunmiseawofadeju200@gmail.com
Get in touchWhatsApp (opens in a new tab)