We dig into hard problems.
That's the whole thing.

Ahris studied Computer Science at IBA Karachi, where the program pushed him toward systems thinking early. He started in deep learning during an apprenticeship at Folio3 AI, working on Generative Adversarial Networks before GANs were a popular thing to put on a portfolio. That research instinct stayed.

At Sofstica Solutions, he moved into backend engineering and cloud infrastructure — building production systems in Spring Boot and FastAPI, working with LangGraph and RAG pipelines, managing AWS deployments. He's one of those engineers who can shift between an ML training loop and a production API without losing the thread of what the system is supposed to do.

His notable projects include an AI clinical copilot for respiratory therapy teams, a full-stack document intelligence platform, and a GAN-based model for synthesizing handwritten Urdu script — a genuinely underserved problem in the NLP space.

Ali studied at LUMS, which gave him a rigorous systems foundation. His professional work since graduation has been almost entirely in the hard end of backend engineering — distributed systems, graph databases, multi-integration data platforms.

At Kaleidoscope, he built the backend of a unified context platform that pulls data from over 20 cloud, DevOps, and security tools and represents it as a graph using Amazon Neptune. He also built the crawler framework that made it reliable: a credential validation layer across 23 crawlers that cut failed runs by 30%, and a LocalStack-based test environment that brought integration test time down from 13 minutes to 2.

Beyond that, he's built an AI agent runtime with real-time streaming in Go, a secure file-sharing platform, a SaaS website builder, and an autonomous trading system. He also has merged contributions in Prometheus, Apache Airflow, and Backstage — which says something about his comfort reading and working in large, unfamiliar codebases.