AI/ML Engineer turning
raw data into working systems.
I'm Vaibhav Khator — building machine learning models, generative AI applications, and analytics dashboards that hold up outside a notebook. Based in Kolkata, India.
Grounded in fundamentals, built for applied AI.
AI/ML Engineer and data professional with a Master of Computer Applications (Artificial Intelligence & Machine Learning) from Chandigarh University and a Bachelor of Computer Applications from IEM, Kolkata.
Skilled in Python, SQL, Machine Learning, Deep Learning, FastAPI, Power BI and Tableau, with hands-on experience building AI-driven applications, data pipelines, and analytics dashboards. I translate complex data into actionable insight and automate workflows to improve efficiency — with a strong foundation in software engineering, NLP, and generative AI.
Where I've applied it.
Oct 2024
Data Management Intern
NIRMAAN- Designed and implemented structured data workflows to organize user records, improving accuracy and consistency across the dataset lifecycle.
- Collected, cleaned, and validated large volumes of user data, reducing inconsistencies and strengthening dataset reliability for downstream use.
- Analyzed real-time data streams to support timely, accurate reporting and decision-making.
- Organized and filtered datasets to enable clear, structured presentation of key metrics to stakeholders.
- Collaborated with cross-functional teams to support personalized service delivery through effective data management practices.
Volunteer, Data Management
Inamigos Foundation- Onboarded and trained new volunteers, ensuring smooth integration into ongoing initiatives.
- Supported awareness and fundraising campaigns by coordinating outreach efforts and campaign materials.
- Collaborated with team members during virtual events to drive engagement and participation.
- Proposed outreach strategy improvements that enhanced cross-team communication and collaboration.
Toolkit, indexed.
Things I've built.
AI Voice Receptionist
An AI-powered virtual receptionist using LLMs, LangChain, Ollama, and RAG on a FastAPI/PostgreSQL backend — with Google Calendar integration for automated appointment booking and prompt engineering to sharpen response accuracy.
Hybrid Multimodal Search Engine
A semantic search engine built with FastAPI, LangChain, Ollama, and a Qdrant vector database, paired with a Next.js/Tailwind CSS frontend and RAG-based retrieval for context-aware results.
Hospital Management Dashboard
An interactive Power BI/DAX dashboard analyzing patient admissions, revenue, and insurance trends to support data-driven hospital administration decisions.
Healthcare Analytics Dashboard
An executive Power BI dashboard covering monthly KPIs, patient demographics, cost, and revenue trends for U.S. Healthcare Dynamics (2019–2020).
Sentiment Analysis using Machine Learning
A Python/Scikit-learn sentiment classification model using TF-IDF and Word2Vec for feature extraction and text preprocessing.
Image Caption Generator
A CNN-based deep learning image captioning system built in Python and deployed through an interactive Streamlit interface.
Weather Forecasting using AutoML
An AutoML pipeline in Python for time-series weather forecasting, with feature engineering to improve model performance.
Milk Production Prediction (ANN)
A TensorFlow-based regression neural network to predict milk production trends, with data preprocessing and model tuning.
Portfolio Website (v1)
A responsive personal portfolio built with HTML, CSS, and JavaScript, with a functional contact form, hosted on GitHub Pages — the site this one replaces.