Document Intelligence Assistant
A retrieval-augmented assistant that answers questions from uploaded PDFs with citations, chunk inspection, semantic search, and confidence-aware responses.
AI Engineer · ML Systems · Applied Research
Building machine intelligence with a clear signal.
I design and build practical AI systems — from machine learning experiments and RAG pipelines to data-driven products that turn messy information into usable decisions.
Selected Work
Projects with working models, not decorative buzzwords.
Visual Defect Classifier
Image classification pipeline for detecting visual anomalies using transfer learning, augmentation, confusion-matrix analysis, and lightweight deployment.
Sentiment & Intent API
A compact NLP service that classifies sentiment and user intent, exposing clean endpoints, logging, validation, and a simple evaluation dashboard.
Time-Series Forecast Lab
A forecasting playground comparing baseline models, feature engineering strategies, cross-validation windows, and experiment tracking for reproducible results.
Technical Stack
Tools I use to move from notebook to product.
Core strengths
- Python / ML
- Data Handling
- Deep Learning
- LLM Apps
- Frontend
Working toolkit
Python
NumPy
Pandas
Scikit-learn
PyTorch
TensorFlow
OpenCV
Transformers
LangChain
Vector Search
FastAPI
Flask
SQL
Git
Docker
MLflow
Streamlit
HTML/CSS/JS
Interests
The areas I keep returning to.
01
Applied LLM Systems
RAG, agents, evaluation, prompt design, and reliable AI workflows.
02
Computer Vision
Classification, detection, segmentation, and real-world image pipelines.
03
Data Products
Dashboards, APIs, and interfaces that make model outputs understandable.
04
Model Evaluation
Metrics, error analysis, reproducibility, and honest measurement.
Blog / Notes
Short notes from experiments, failures, and rebuilds.
What makes a RAG answer trustworthy?
Notes on citations, retrieval quality, chunking, and why confident language can hide weak context.
My model failed. The dataset explained why.
A practical debugging checklist for noisy labels, leakage, imbalance, and misleading accuracy.
From notebook to small API.
How I wrap experiments into endpoints with validation, logging, and a deployment-friendly structure.
Let’s build something that learns clearly.
I’m interested in AI engineering roles, ML projects, research collaborations, and product ideas where machine learning needs to be useful, measured, and shipped.