A Little Less Conversation, a Little More Action, Please
Investigating the Physical Common-Sense of LLMs in a 3D Embodied Environment
About
I'm Muhammad Maaz Akram, a bs artificial intelligence @ air university - 6th semester based in Islamabad, Pakistan. I specialize in building automation systems that connect platforms, streamline workflows, and solve real business problems using AI and modern integration tools.
My work sits at the intersection of artificial intelligence and practical automation — from AI-powered content generation pipelines to multi-channel outreach systems built with n8n and GoHighLevel. I approach every problem by reading documentation thoroughly, understanding the APIs involved, and building systems that are reliable and scalable.
When I'm not building automations, I study research papers on LLMs, transformer architectures, and embodied AI to deepen my understanding of the systems I work with.
Education
BS Artificial Intelligence
Air University, Islamabad
Focus Areas
Workflow Automation · AI Systems
API Integration · LLM Research
Capabilities
Technologies and platforms I work with regularly.
Workflow and system automation platforms
Core languages and frameworks
Research and applied machine learning
Applied data, signal, and vision workflows
Integration and infrastructure
Marketing and business tools
Research
Key research papers that shape how I think about AI and systems.
Investigating the Physical Common-Sense of LLMs in a 3D Embodied Environment
The Transformer architecture that changed NLP
Exploring recursive structures in language modeling
Live Activity
GitHub Contributions
@MaazAkram76
Loading latest push…
Duolingo Streak
…
@MMaazAkram
Portfolio Evidence
12 projects
Automation, NLP, CV, biomedical AI, and full-stack systems
Currently
Building
AI × Automation · Islamabad, Pakistan