We're a small, senior AI engineering team based in Hyderabad. No padding, no juniors, no offshore bench. Just four people who've spent their careers at the intersection of data, AI, and software — building things that have to work in production.
Drut AI was started because we kept seeing the same problem: impressive AI demos that crumbled under real-world data, latency requirements, and edge cases. We've all worked in large organisations where AI projects died in pilot. We built Drut AI to fix that.
Our team combines deep research credentials with hands-on engineering. We've built ML systems for banks, hospitals, and logistics firms. We've published research on retrieval and memory. We've written the infrastructure that those systems run on.
We take a small number of engagements at a time — because we believe that the only way to do this well is to go deep. Every client we work with gets the full team, not an account manager and a junior.
Builds the systems that make the research actually ship. From agent orchestration frameworks to fine-tuning pipelines, He turns experimental AI into production-grade software. Has a habit of rewriting things in Rust when Python gets too slow.
The scientific backbone of Drut AI. Drives research on memory architectures, retrieval systems, and model evaluation. Published work spans NLP, information retrieval, and scalable ML systems. The kind of person who reads papers on a Friday night.
Translates business problems into AI architectures. SK has designed and deployed AI systems for enterprises across BFSI, logistics, and healthcare. Obsessed with making AI reliable, explainable, and cost-efficient in production.
Every small team has one person who remembers the deadlines, books the calls, keeps the humans talking to each other, and somehow makes it all feel effortless. At Drut AI, that's Shyam. New to the team, not new to the work.
We read the papers, run the evals, and measure everything. We don't ship demos that can't survive production. Every claim we make has a number behind it.
If we build something useful that doesn't expose our clients, we open-source it. The AI ecosystem got us here — we owe it contributions, not just consumption.
We don't disappear after deployment. Our best work happens when clients bring us deep domain knowledge and we bring the AI engineering. A true exchange.
Four people with 40 years of combined experience. No junior bench. No account managers. When you work with Drut AI, you work with the people who built it.
We take on a small number of new clients each quarter. If you have a serious AI problem and want a senior team to solve it, let's talk.