I build scalable AI applications that combine LLMs, autonomous agents, RAG pipelines, and intelligent chatbots. My work bridges cutting-edge AI research with production-ready systems, delivering real-world impact.
- 🧠 Domains: LLMs, NLP, Computer Vision, Time-Series Analysis, Generative AI, Reinforcement Learning, Autonomous Agents
- ⚡ Tools: Python, PyTorch, TensorFlow, Hugging Face Transformers, LangChain, OpenAI API, FastAPI, Django, FAISS, Chroma
| Project | Description | Tech / Skills |
|---|---|---|
| AQI Prediction using Time-Series Analysis | Forecasted future air quality trends in Islamabad using state-of-the-art forecasting models | Python, Pandas, scikit-learn, Time-Series Analysis, ML |
| Atari Games Reinforcement Learning | Trained agents to play Atari games using Deep Q-Learning in OpenAI Gym environment | Python, PyTorch, OpenAI Gym, Reinforcement Learning |
| Pricegram - E-commerce Chatbot | Built an e-commerce chatbot using Llama 3.1 & LangChain, incorporating RAG pipelines and NL2SQL, powered by custom-scraped data | Python, LangChain, LLMs, RAG, Back-End Development, Vector Databases |
| Speaker Diarization | Trained a speaker diarization model using Segmentation 3.1 to segment & identify speakers in audio | Python, Machine Learning Algorithms, Audio Processing |
| Traffic Detection and Tracking | Utilized YOLOv8 to detect and track vehicles, counting city traffic accurately | Python, YOLOv8, Computer Vision, Object Detection |
| Drone Detection using YOLOv11 | Fine-tuned a YOLOv11-nano model on custom drone dataset for real-time detection | Python, PyTorch, Ultralytics, Object Detection |
| AVATARé AI Agents | Built a platform for orchestrating organizational AI agents | Python, LangChain, RAG, Agent Orchestration |
| AVATARé Land Banker | Developed backend for land bankers & developers collaboration platform | Python, FastAPI, Vector Databases |
| Custom n8n NexorOne Node | Developed a custom n8n integration node for the NexorOne API, an internal payment processing platform used by AVATARé for workflow automation and transaction orchestration | TypeScript, n8n, REST APIs, Automation, Backend Development |
| SettledIQ Autoglass | Built and contributed to a platform connecting customers, autoglass shops, and insurance advocates, streamlining claim management and operational workflows | FastAPI, Python, MySQL, Backend Systems, API Development |
| Time-Series Forecasting Experiments | Worked on multiple datasets for stock price & sensor data prediction using classical & neural models | Python, scikit-learn, LSTM, Prophet, ML |
| ML Ops / Deployment Pipelines | Deployed ML models in production with FastAPI, Docker, and vector database integration | Python, FastAPI, Docker, Vector DBs, ML Ops |
Languages: Python | C++ | Java
Frameworks & Libraries: PyTorch | TensorFlow | scikit-learn | Hugging Face Transformers | LangChain | FastAPI | Django | OpenAI API
Databases & Others: MySQL | MongoDB | Vector Databases (FAISS, Chroma) | GraphQL | Postman | n8n
LinkedIn | GitHub | afaisal.bese21seecs@seecs.edu.pk
