I'm Adekunle AbdulMuheez, an AI & Software Engineer focused on building reliable, production-grade systems around AI components.

My engineering work sits at the intersection of:

Artificial Intelligence × Software Engineering × Data × Product

I specialize in building systems around:

• 🧠 LLM foundations and applications
• 🔎 Retrieval-Augmented Generation (RAG)
• 🗄️ Vector search and grounding
• 🤖 Agentic systems and AI workflows
• ⚡ High-performance backend infrastructure
• 🏗️ Multi-tenant SaaS architecture
• 🏥 AI-powered healthcare systems
• 📊 AI analytics and data applications

I care about more than making an AI demo work.

I want to build systems that can be tested, evaluated, improved, deployed, and operated reliably in production.

My development process follows a continuous improvement loop:

        ┌──────────────┐
        │    BUILD     │
        └──────┬───────┘
               ↓
        ┌──────────────┐
        │   EVALUATE   │
        └──────┬───────┘
               ↓
        ┌──────────────┐
        │   ANALYZE    │
        └──────┬───────┘
               ↓
        ┌──────────────┐
        │   IMPROVE    │
        └──────┬───────┘
               ↓
        ┌──────────────┐
        │   DEPLOY     │
        └──────┬───────┘
               │
               └──────────────→ Repeat

Build it. Measure it. Understand it. Improve it. Ship it.

My focus is moving beyond simply consuming AI APIs toward understanding how complete AI-powered systems are engineered.

LLM ENGINEERING
│
├── LLM Foundations
├── Prompt Engineering
├── Structured Outputs
├── Tool / Function Calling
├── Context Engineering
└── Model Integration

GROUNDING & RETRIEVAL
│
├── RAG Architectures
├── Embeddings
├── Vector Search
├── ChromaDB
├── FAISS
└── Retrieval Pipelines

AGENTIC SYSTEMS
│
├── AI Agents
├── LangChain
├── LlamaIndex
├── LangGraph
├── Tool Use
└── Multi-Step Workflows

PRODUCTION AI
│
├── Evaluation
├── Reliability
├── Observability
├── Latency
├── Cost Optimization
└── Deployment

Also: PyTorch • TensorFlow • Keras • Ollama

Core technologies

• Python
• FastAPI
• SQLModel
• SQLAlchemy
• Alembic
• Pydantic
• Node.js
• Express
• C++
• Java
• RESTful APIs

Web

• TypeScript
• JavaScript
• Next.js
• React
• Svelte
• Streamlit
• Tailwind CSS
• Framer Motion
• Ant Design

Mobile

• React Native
• Android
• Kotlin
• Jetpack Compose
• CameraX

Databases

• PostgreSQL
• Neon
• Supabase
• Firebase
• Redis

Infrastructure & Development

• Docker
• Vercel
• Render
• Railway
• Git
• GitHub
• Linux
• WSL2

Native Android AI debugging assistant

DevLens bridges the gap between physical screens and instant debugging.

It captures code or terminal error output using CameraX + OCR, then uses the Google Gemini API to analyze the captured content and provide line-by-line diagnostics and potential fixes through a Jetpack Compose interface.

Android
├── Kotlin
├── Jetpack Compose
├── CameraX
└── OCR

AI
└── Google Gemini API

Physical Screen
      ↓
CameraX
      ↓
OCR
      ↓
Captured Code / Error
      ↓
Gemini
      ↓
Analysis
      ↓
Line-by-Line Diagnosis
      ↓
Suggested Fix

MedOS is an AI-augmented, multi-tenant hospital management platform designed to bring healthcare operations, intelligent assistance, and secure data management into one system.

The long-term goal is to evolve MedOS beyond a conventional hospital dashboard into a production-grade healthcare SaaS platform.

                         MEDOS
                           │
             ┌─────────────┴─────────────┐
             │                           │
             ▼                           ▼
      APPLICATION LAYER              AI LAYER
             │                           │
      ┌──────┴──────┐             ┌──────┴──────┐
      │             │             │             │
   Next.js       APIs           LLMs           RAG
      │             │             │             │
      └──────┬──────┘             └──────┬──────┘
             │                           │
             └─────────────┬─────────────┘
                           ▼
                     PostgreSQL
                           │
                    Tenant Isolation
                           │
             ┌─────────────┼─────────────┐
             ▼             ▼             ▼
          Patients     Operations      Finance

• Multi-tenant architecture
• Database-level tenant isolation
• Role-based access control
• Secure healthcare data workflows
• AI-assisted clinical workflows
• RAG-powered information retrieval
• Inventory management
• Financial record management
• Patient-facing experiences
• Production-ready infrastructure
• Nigerian healthcare considerations
• NDPR-conscious data handling

Vision: Build healthcare software where AI is integrated into the architecture and workflows rather than treated as a standalone feature.

OgaMetrics is an AI analytics platform designed to help small businesses understand their operational data and turn it into actionable insights.

The system combines structured data processing, analytics, and AI-powered retrieval.

Business Data
      ↓
Data Processing
      ↓
Schema Detection
      ↓
Normalization
      ↓
PostgreSQL / Neon
      ↓
Embeddings
      ↓
Vector Search
      ↓
AI Retrieval
      ↓
Operational Insights

• Python
• FastAPI
• PostgreSQL
• Neon
• Clerk
• RAG
• Embeddings
• Vector Search
• AI Analytics

The architecture separates authentication from application data while using serverless PostgreSQL infrastructure for scalable storage.

get2learn is a cloud-hosted learning platform designed to help developers, students, and young technical professionals discover, organize, and track high-quality educational content.

• 🎥 Video discovery
• 🔎 Search and filtering
• 💾 Saved videos
• 📚 Playlists
• ❤️ Reactions
• 📈 Learning progress
• 📝 Private notes
• 🔗 Curator link submission
• 💬 Feedback

Frontend
   ↓
Responsive Web Interface
   ↓
FastAPI
   ↓
PostgreSQL
   ↓
Authentication / RBAC
   ↓
AI-Ready Retrieval Layer

I'm also building ADEXTECH HUB as a technology and software-building initiative focused on creating useful digital products and AI-powered systems.

• AI Engineering
• Software Engineering
• SaaS
• Healthcare Technology
• Business Intelligence
• Web Applications
• AI-powered Products
• Technical Innovation

RankBloom is another product initiative focused on building digital solutions around web presence, discoverability, and growth.

The broader objective is to combine software, automation, data, and AI to solve practical business problems.

• Azure AI Engineer Associate
• Prompt Engineering & Generative AI Specialist
• Azure Cognitive Services Specialist

• Full-Stack Software Developer Professional

• Front-End Development
• Search Engine Optimization (SEO)
• Web Analytics

• Multi-Cloud Solutions Architecture
• AWS
• Azure

My development path is centered around six major areas:

┌────────────────────────────────────────────────────┐
│              AI ENGINEERING                        │
├────────────────────────────────────────────────────┤
│                                                    │
│  01. LLM FOUNDATIONS                               │
│      Tokens • Context • Parameters • APIs          │
│                                                    │
│  02. GROUNDING & RETRIEVAL                         │
│      RAG • Embeddings • Vector Search              │
│                                                    │
│  03. AGENTIC SYSTEMS                               │
│      Tools • Agents • Workflows • MCP              │
│                                                    │
│  04. EVALUATION                                    │
│      Evals • Tracing • Error Analysis               │
│                                                    │
│  05. PRODUCTION                                    │
│      Deployment • Monitoring • Cost • Latency      │
│                                                    │
│  06. ML FOUNDATIONS                                │
│      Data • Bias/Variance • Model Evaluation       │
│                                                    │
└────────────────────────────────────────────────────┘

AI ENGINEERING
      │
      ├── LLM Applications
      ├── RAG Systems
      ├── AI Agents
      ├── Tool Calling
      ├── Context Engineering
      └── AI Evaluation
             │
             ▼
SOFTWARE ENGINEERING
      │
      ├── Backend Architecture
      ├── APIs
      ├── Databases
      ├── Multi-Tenant SaaS
      ├── Infrastructure
      └── System Design
             │
             ▼
REAL-WORLD PRODUCTS
      │
      ├── Healthcare
      ├── Analytics
      ├── Developer Tools
      └── AI SaaS

I think about AI applications as engineering systems, not just prompts.

User Problem
     ↓
System Design
     ↓
Data / Context
     ↓
Retrieval
     ↓
Model
     ↓
Tools / Actions
     ↓
Evaluation
     ↓
Observability
     ↓
Deployment
     ↓
Continuous Improvement

The objective is not simply:

"Can the model answer?"

It is:

"Can the entire system produce useful, reliable, measurable results?"

I'm particularly interested in:

• 🤖 AI Engineering
• 🧠 LLM applications
• 🔎 RAG and retrieval systems
• 🧩 Agentic architectures
• ⚡ Backend infrastructure
• 🏗️ Distributed and scalable systems
• 🏥 Healthcare technology
• 📊 AI analytics
• ☁️ Cloud architecture
• 🔐 Secure multi-tenant SaaS
• 🌐 Open-source AI tooling

• Build full-stack applications
• Build AI-powered applications
• Work with LLM APIs
• Explore RAG architectures
• Build multi-tenant application concepts
• Build production-grade AI systems
• Deepen AI evaluation practices
• Build reliable AI agents
• Improve system design expertise
• Deploy scalable SaaS products
• Contribute more to open source
• Build technology solving real-world problems

My preferred learning loop is simple:

LEARN
  ↓
BUILD
  ↓
BREAK
  ↓
DEBUG
  ↓
UNDERSTAND
  ↓
IMPROVE
  ↓
DEPLOY
  ↓
REPEAT

I believe the fastest way to understand a technology deeply is to use it to solve an actual problem.

I'm open to conversations and collaboration around:

• Open-source AI tooling
• AI agents
• RAG systems
• AI SaaS
• Backend architecture
• Healthcare technology
• Developer tools
• High-scalability infrastructure
• Interesting technical problems

Building intelligent systems that solve real problems.