Forever learner of AI, I'm passionate about both - its ability to automate labor-intensive tasks and its deeper mission of modeling human intelligence. I'm currently building AI assistants for Cisco Security products using LLMs, after completing my MS in AI at CMU.
Previously, I built products from scratch as a software engineer at McKinsey, and now I'm focused on GenAI and multimodal ML (voice-vision-language). I aspire to create AI-driven products that empower people and push the boundaries of intelligence—human and artificial.
I'm an AI researcher at Cisco developing LLM-powered assistants that help users troubleshoot security products. With a Master's from Carnegie Mellon's School of Computer Science, I specialize in turning cutting-edge AI research into practical solutions.
My experience spans from innovative startups to Fortune 500 companies, where I've consistently delivered scalable AI systems that solve real-world problems. I thrive at the intersection of research and application, building bridges between theoretical breakthroughs and production-ready solutions.
LLMs, RAG, Multimodal ML
NLP, Transformers, Computer Vision
ML Operations, Scalable Pipelines
Few-shot Learning, Generative AI
Demonstrating career progression and impact across diverse organizations
Cisco
Developing LLM assistants with RAG and NL-to-SQL. Awarded strategic recognition (top 20% of contributors).
Faire
Automated credit approvals for 30% of applicants, improving accuracy by 5%.
McKinsey
Built code auto-generation tools, revamped MedTech client website, reduced hiring costs by 90%.
Toshiba
Delivered few-shot image classification achieving 96% accuracy.
Blue Sail
Created Android app for visually impaired with 98% object detection accuracy.
Showcasing technical capabilities through real-world applications and research contributions
Advanced research project improving CodeLlama with library documentation integration using Retrieval-Augmented Generation techniques.
Generative AI solution for SGK creating brand-specific product backgrounds using advanced diffusion models and computer vision.
Multimodal machine learning research project creating expressive 3D face animations from voice input using deep learning.
Award-winning hackathon project using BLIP-2 and LLM prompt engineering to provide intelligent outfit recommendations and style analysis.
ACM COMPASS, 2025
Our work focuses on unlocking historical housing data using OCR and machine learning to better understand the long-term impacts of policies like redlining. By digitizing and predicting property values from the 1930s, we provide tools for social science researchers to analyze how these policies shaped wealth and inequality.
I'm always excited to discuss AI research, machine learning challenges, or potential collaborations. Whether you're looking for an AI researcher, ML engineer, or technical consultant, let's explore how we can create intelligent solutions.