I'm an AI engineer and founder from Bangladesh, focused on building intelligent systems from the ground up.

My interests sit at the intersection of Artificial Intelligence, Large Language Models, Agentic AI, Cybersecurity, and AI Systems Engineering.

I enjoy going beyond simply using AI models — I want to understand how they work, how they are trained, and how complete AI systems can be engineered around them.

I'm currently exploring and building systems around:

• 🤖 Large Language Models (LLMs)
• 🧠 Agentic AI & Multi-Agent Systems
• 🔬 AI/ML Research
• ⚡ AI Systems & Model Architecture
• 🛡️ AI for Cybersecurity
• 💻 Agentic Coding Systems
• 🔎 Retrieval-Augmented Generation (RAG)
• 🧩 Memory Systems for AI Agents
• 🏗️ Distributed & Decentralized AI
• 🐍 Python & PyTorch
• 🦀 Rust for high-performance AI tooling

My long-term goal is to contribute to the development of frontier-level AI systems and build technology from first principles rather than simply depending on existing solutions.

Craftly is an AI-focused technology company I'm helping build with a long-term vision of creating powerful AI products and infrastructure.

One of the projects I'm working around is Craftly Robot — an AI system designed to move beyond simple chatbot interactions and become capable of performing complex software-building tasks.

I'm also building Craftly Workspace, an internal platform designed around:

• Organization management
• Onboarding
• Secure identity & access
• Organizational hierarchy
• Task management
• Evidence-based operations
• Internal communication
• Resources & training
• Team collaboration

The long-term vision is to build an ecosystem where people and intelligent systems can work together efficiently.

I'm also involved in building Aeitron, a frontier-intelligence focused project exploring areas such as:

• Agentic AI
• Cybersecurity
• Intelligent coding systems
• LLMs
• AI agents
• Autonomous reasoning
• AI infrastructure

I'm particularly interested in building systems that can understand complex repositories, reason about software, execute tasks, generate changes, and verify their own work.

Some of the areas I'm actively studying and experimenting with:

• Transformer architectures
• Tokenization
• Embeddings
• Attention mechanisms
• Positional encoding
• Pretraining
• Fine-tuning
• Model evaluation
• Knowledge distillation
• Mixture-of-Experts (MoE)

• Multi-agent architectures
• Tool use
• Planning
• Task graphs
• Agent memory
• Context management
• Autonomous coding
• Verification loops
• Agent orchestration

• Model serving
• Distributed training
• GPU computing
• Data pipelines
• Inference optimization
• Memory systems
• Evaluation infrastructure
• AI security & guardrails

I'm particularly interested in the intersection of:

AI × Cybersecurity × Autonomous Agents

and how intelligent systems can help analyze, defend, and reason about complex software environments.

I don't want to treat AI as just an API.

I want to understand the entire stack:

Data
 ↓
Data Quality & Filtering
 ↓
Tokenization
 ↓
Model Architecture
 ↓
Pretraining
 ↓
Post-training
 ↓
Evaluation
 ↓
Inference
 ↓
Agents
 ↓
Tools
 ↓
Memory
 ↓
Verification
 ↓
Production AI System

That's why I'm continuously learning both the mathematics behind machine learning and the engineering required to turn models into real systems.

• Python
• PyTorch
• Machine Learning
• Deep Learning
• LLMs
• RAG
• Agentic AI
• Neural Networks

• Rust
• Node.js
• FastAPI
• REST APIs
• Distributed Systems
• Backend Architecture

• CUDA
• GPU Computing
• Distributed Training
• FSDP
• DeepSpeed
• Model Serving
• Docker
• Kubernetes
• PostgreSQL
• Redis
• S3 / MinIO

• Flutter
• Node.js
• Python
• TypeScript
• Full-stack application development

I'm especially interested in building complete AI platforms composed of multiple specialized systems.

A typical architecture I'm exploring looks like:

                    ┌────────────────────┐
                    │     AI Gateway     │
                    └─────────┬──────────┘
                              │
                    ┌─────────▼──────────┐
                    │ Identity & Access  │
                    └─────────┬──────────┘
                              │
              ┌───────────────▼───────────────┐
              │        Model Serving          │
              └───────────────┬───────────────┘
                              │
        ┌─────────────────────▼─────────────────────┐
        │              Agent Runtime                │
        └───────────────┬───────────┬──────────────┘
                        │           │
                ┌───────▼───┐ ┌─────▼─────┐
                │   Tools   │ │  Memory   │
                └───────┬───┘ └─────┬─────┘
                        │           │
                        └─────┬─────┘
                              │
                    ┌─────────▼─────────┐
                    │ Verification      │
                    │ & Evaluation      │
                    └───────────────────┘

I'm fascinated by the engineering challenges involved in making these systems reliable, scalable, secure, and autonomous.

I'm continuously strengthening my foundations in:

• Mathematics for Machine Learning
• Linear Algebra
• Probability & Statistics
• Calculus
• Optimization
• Deep Learning
• Transformer architectures
• LLM training
• Distributed systems
• AI infrastructure
• Cybersecurity
• Systems programming with Rust

I believe strong AI engineering requires both mathematical foundations and serious software engineering.

Some of the areas I work on across my repositories include:

• AI Engineering systems
• Agentic AI
• LLM experimentation
• RAG systems
• Memory architectures
• Cybersecurity AI
• AI finance applications
• Full-stack applications
• Internal business infrastructure
• Autonomous coding systems

Check out my repositories below to see what I'm currently building and experimenting with.

I'm actively using GitHub to document my learning, experiments, research, and engineering work.

I care about:

• Building consistently
• Open-source experimentation
• Writing maintainable systems
• Learning from failures
• Iterating quickly
• Sharing what I learn

Build. Break. Understand. Rebuild.

My long-term goal is to become a world-class AI engineer and researcher capable of building frontier AI systems from first principles.

I want to work on problems that are difficult today — especially around:

Intelligence → Agents → Reasoning → Autonomous Software Engineering → AI Security

And ultimately, I want to build technology that can have a meaningful impact far beyond a single project.

I'm from Bangladesh, and I believe talented people shouldn't need to leave their home country to build ambitious technology.

One of my long-term ambitions is to help create an environment where young engineers and researchers from Bangladesh can participate in building globally competitive technology.

I'm always interested in meeting people who are passionate about:

• Artificial Intelligence
• Machine Learning
• LLMs
• Agentic AI
• Cybersecurity
• Open Source
• Research
• Startups
• Software Engineering

GitHub: @iammahmudhasan

Instagram: @iammahmudhasan

LinkedIn: @iammahmudhasan

X: @iammahmudhasan

• 🇧🇩 Building from Bangladesh
• 🤖 Obsessed with AI systems
• 🧠 Learning how intelligence can be engineered
• 🛠️ Building instead of just consuming
• 🔬 Interested in research and first-principles thinking
• 🚀 Founder mindset
• 💻 Engineer mindset
• 📚 Always learning

⭐ If you find something interesting in my repositories, feel free to explore, star, or contribute.