I'm a final-year BS Artificial Intelligence student at COMSATS University Islamabad, Wah Campus, and Chief AI Officer (CAIO) at Algoligence, an AI engineering startup. I also serve as Community Manager & Founding Member of Google Developer Group (GDG) On Campus Wah, where I help run workshops and mentor juniors in the campus AI/DS community.
I don't just study AI, I build with it. My work spans full-stack AI applications, computer vision systems, agentic and multi-agent workflows, and sequence models. I'm actively looking for remote roles, project collaborations, and open-source contributions where I can bring real ML engineering skills to the table.
adeel = {
"role" : "CAIO @ Algoligence",
"currently_building": ["Agentic AI systems", "Multi-agent architectures", "LangGraph workflows"],
"open_to" : ["Remote AI/ML roles", "Collabs", "Research opportunities"],
}💻 Languages
🧠 Deep Learning & ML
🤖 Agentic AI & Backend
📱 App & Web
📊 Data Science
⚙️ DevOps & Tools
|
Multi-agent GraphRAG platform that ingests a repo's full git history, PRs, and AST to explain why code is the way it is, forecast blast radius before a change, and propose refactors through an adversarial Proposer/Critic debate gated by human review.
|
Four LangGraph agents (Clarity, Research, Validator, Synthesis) collaborate on business research queries, pausing to ask the user when a query is ambiguous and looping back for more evidence when confidence runs low.
|
|
AI content capture platform that eliminates the "digital graveyard" problem. Gemini Vision for screenshot OCR, Groq Whisper for video transcription, real-time SSE updates, PWA support.
|
Geospatial similarity engine built on Google Earth Engine's satellite embeddings (DeepMind AlphaEarth Foundations). Pick a reference location and it ranks and visualizes environmentally similar sites across a region, no hand-engineered indices needed.
|
|
Real-time British Sign Language classifier across 34 classes (alphabet + numbers). MobileNetV2 with MediaPipe hand landmark preprocessing on 34,000 images.
|
Malware binary classifier using the byte-to-image technique (Nataraj et al., 2011). ResNet50 trained on Malimg across 25 malware families. Full-stack security ops dashboard.
|
|
AI-powered skin lesion classifier across 7 diagnostic classes using MobileNetV2 on HAM10000. Full three-tier architecture with JWT auth, scan history, and risk assessment UI.
|
ETL pipeline and Tableau command-center dashboard turning 500,000+ UK retail transactions into inventory, seasonality, and expansion insights, cleaned and modeled end to end in Python.
|
- 🎓 Machine Learning Specialization (DeepLearning.AI / Coursera)
- 🎓 IBM Data Analysis with Python (IBM / Credly)
- 🎓 CS50P: Introduction to Python (Harvard University)
I'm open to remote AI/ML roles, agentic AI & applied ML collaborations, and research opportunities, across computer vision, deep learning, and multi-agent systems. Let's work together.