Skills & Expertise

The tools I use to build and ship AI products, each backed by a project

LLMs & AI Engineering

LLM Applications

OpenAI API (GPT-4o-mini), prompt design, intent extraction and context management. Used in BrewHub.

Voice AI

Whisper speech-to-text, ElevenLabs and gTTS text-to-speech, real-time browser audio pipelines.

Conversation State

Finite-state machines for slot filling, which keep LLM dialogues complete and predictable.

NLP

Text processing, skill and entity extraction, intent classification, TF-IDF similarity.

Machine Learning & Data

Python

Main language for ML pipelines, data processing, APIs and automation.

scikit-learn

Regression, classification, clustering, vectorisation, model tuning and pipelines.

pandas, NumPy & SciPy

Data cleaning, merging multi-source datasets, de-duplication and sparse matrices.

Feature Engineering & EDA

Exploratory analysis and feature design. Studied data mining at NPTEL, IIT Kharagpur.

SQL / MySQL

Schema design, complex queries and data modelling.

Tableau

Interactive dashboards and BI reporting.

Backend, Cloud & Delivery

FastAPI

REST services for LLM and voice AI backends, including the BrewHub API.

Google Cloud Platform

Deploying containerised services on Cloud Run.

API Integration

OpenAI, ElevenLabs, Apollo.io and SMTP, with fallbacks and usage tracking.

React & JavaScript

React + Vite + Tailwind frontends for AI products.

Streamlit

Fast interactive apps and dashboards for ML models.

Git, GitHub & Linux

Version control, the command line and shell scripting.

Currently Building With

LangChain & LlamaIndex

LLM orchestration and retrieval-augmented generation (RAG) workflows.

Vector Databases

Pinecone and Chroma for embeddings and semantic search.

AI Workflow Automation

n8n, VAPI, webhooks, async API handling and JSON orchestration.

Also Worked With

FlutterHTML / CSSTailwind CSSFigmaSoftware Testing