An AI outbound engine — a dedicated BDR team, working 24/7. Watches the internet for real buying signals, writes outreach tuned to each lead's pain points, and books 5–7 qualified meetings a week straight into customers' calendars. Live in production with paying pilots and outcome-based pricing — customers only pay for meetings booked.
End-to-end AI product engineering for 2M weekly active users. Built RAG pipelines, custom models, and image embedding infrastructure. Took over backend engineering for the US team rollout. Did everything from product specs and AI experiments to shipping in the app.
Helped scale the tech organization at a growing IoT unicorn. Saw firsthand how scaleup ops run, and how much of it could be automated. Validated the conviction behind fifteenA.
Built the AWS infrastructure (Batch, S3, Lambda, Aurora) powering an ML-driven digital twin for a DAX40 client. End-to-end ownership from Palantir Foundry data pipelines to production deployment.
First employee at the LLM observability startup. Shipped full-stack features (Next.js, React, tRPC) during the product-market fit phase. Now 20K+ GitHub stars, 26M+ monthly SDK installs.
AI-powered workforce platform automating HR, customer support, and analytics via Slack and email integration. Built with Mustafa in 48 hours.
Computer vision system for visually impaired navigation. Uses Mast3r-SLAM, SAM, YOLO, and BLIP for dense 3D reconstruction, collision detection, instance segmentation, and object tracking to provide real-time auditory guidance for obstacle avoidance.
AI app that generates personalized comic books where kids are the main character in their own story. Built it for my nephew, now his friends love it too.
Deep dives into computational photography from Berkeley's CS 180. Face morphing, seamless image blending, panoramic mosaics built from scratch, and neural radiance fields , all implemented from the ground up.