Currently crafting intelligent systems at Futurepath AI , where I've built everything from voice assistants with 97% precision to AI agents querying Kubernetes pods — all with real-time LLM observability dashboards.
Previously at fold.money, I developed scalable natural language tools to help users make sense of their transactions — basically making money talk.
I've worked across multiple startups, collaborating with like-minded engineers, building practical AI systems, and pushing the limits of what's possible with language models.
When I'm not in front of a computer screen, I'm probably in front of my phone binging. Time for a lifestyle upgrade 😅
Worked in a small team to create the beautiful tagging engine with F1 score of 95%!
Spoke at length about scaling RAG for Large Number of Records
A website which based on symptoms entered by user gives the probable disease he/she may have and recommends nearby hospitals/clinics (within 10 KM).
This webapp lets user to listen to songs based on realtime emotion detected through mobile camera or web camera and helps our users to interactively chat with similar users matched on the emotion they currently feel.
Chrome extension to highlight relevant items based on your search query.
Built a search and chat tool with Qdrant and Meilisearch to enhance retrieval of bookmarked articles.
Web-based inference using ONNX Runtime and WebAssembly for performing sentiment analysis directly in browsers.
WhatsApp-based chatbot leveraging Pinecone, Langchain, and GPT for personalized vector-based responses.
Used transfer learning with VGG to classify infected cells, deployed with Streamlit and Docker on Google Cloud Run.
Trained InceptionV3 model via transfer learning to classify 26 alphabetic sign language gestures.