ShopAgent
FlagshipCurrently WorkingAn agent-based shopping workflow designed to simplify product discovery and booking.
AgentsBooking AgentHITL
Hi, I'm
BAHAWAL KHAN
AI / ML ENGINEER
I am an AI/ML Engineer. I build Machine Learning systems. I work with Deep Learning, NLP & Transformers. I build RAG applications. I develop Agentic AI systems.
Building intelligent systems that solve real-world problems.
About
I build AI systems, not just models.
I'm Bahawal Khan, a BS Computer Science student at the University of Lahore (UOL), currently in my 7th semester with a CGPA of 3.06. My focus is on Artificial Intelligence and Machine Learning — spanning classical machine learning, deep learning, NLP, transformers, Generative AI and Agentic AI.
3.06
CGPA
7th
Semester
BS CS
Program
Alongside AI/ML, I also have hands-on web development experience, building interfaces and APIs with HTML, CSS, JavaScript, React.js, Next.js and FastAPI. I enjoy turning ideas into practical, end-to-end systems that can retrieve information, use tools, maintain context and interact with users.
Python
Machine Learning
Deep Learning
NLP
Transformers
Generative AI
RAG
Agentic AI
Capabilities
Technical Skills
Grouped by discipline — from classical ML through deep learning, generative and agentic AI, to the retrieval and tooling that ties it together.
Machine Learning
(11)EDA
Feature Engineering
Linear Regression
Logistic Regression
K‑Nearest Neighbors (KNN)
Random Forest
XGBoost
Model Evaluation
Hyperparameter Tuning
SHAP
Optuna
Deep Learning
(10)ANN
CNN
RNN
LSTM
GRU
Transformers
Self-Attention
DistilBERT
EfficientNetB0
Transfer Learning
NLP
(6)NLP
TF-IDF
Text Classification
Embeddings
Semantic Search
Transformer-based NLP
Generative AI
(11)LLMs
RAG
LangChain
LangGraph
Prompt Engineering
AI Agents
Tool Calling
Memory
Context Management
Hugging Face
Model APIs
Agentic AI
(6)Agents
7 Tools
Booking Agent
HITL
Tool Calling
LangGraph
Database & Retrieval
(6)PostgreSQL
pgvector
FAISS
Chroma
SQL
Vector Databases
AI / Developer Tools
(9)Groq
Hugging Face
LangSmith
Cloudflare
Git
GitHub
Docker
FastAPI
External APIs
Multimodal AI
(4)Vision / Image-to-Text
PDF QA
Voice
Image Generation
Web Development
(5)HTML
CSS
JavaScript
React.js
Next.js
Portfolio
Selected Work
Five systems, one flagship — real architectures, real metrics, and how each one actually fits together.
ChatModel
A multimodal agentic AI assistant capable of interacting with users, retrieving information, using external tools and working across text, vision, PDF, voice and image-generation workflows.
LangGraphRAGPostgreSQLTool Calling
Customer Assistant
BuiltAn NLP/RAG support system that classifies incoming queries and retrieves grounded answers — transformer-based classification paired with embedding retrieval over a Postgres/pgvector-oriented store.
DistilBERTRAGEmbeddingspgvector
PlantCare AI
BuiltA deep learning model fine-tuned to detect plant diseases from images across the PlantVillage dataset — a real agricultural use case, trained end-to-end on CPU.
EfficientNetB0Transfer LearningPlantVillage Dataset
ChurnAI
BuiltAn end-to-end machine learning pipeline predicting customer churn, comparing multiple models before selecting a winner on real test data.
Logistic RegressionRandom ForestHistGradientBoostingClassifierANN
Technical Journey
How the skillset was built
Stage 01
Classical ML
EDA, feature engineering, regression, ensemble models
Stage 02
Deep Learning
ANN, CNN, RNN, LSTM, GRU, transfer learning
Stage 03
NLP
TF-IDF, text classification, embeddings, semantic search
Stage 04
Transformers
Self-attention, DistilBERT, transformer-based NLP
Stage 05
Generative AI
LLMs, LangChain, prompt engineering, model APIs
Stage 06
RAG
Retrieval pipelines, vector databases, context management
Stage 07Current
Agentic AI
LangGraph agents, tool calling, memory, HITL
Engineering
How the technologies connect
An explanatory view of how the stack layers together — not a claim that every project uses every piece.
AI / ML
Classical MLDeep LearningNLP
Generative AI
LLMsRAGLangChainLangGraph
Agentic AI
ToolsMemoryHITLExternal APIs
PostgreSQL / Vector Retrieval
Contact
Let's build something intelligent.
Have an AI project, idea, or opportunity?