Dream11 · engineering brief Speed angle · 2026

Their world

Dream11 rebuilt its entire product surface in one cycle.

When Parliament's online-gaming Act took paid contests off the table, the engineering question changed overnight: how do you move 250 million people onto a free-to-play, second-screen sports entertainment experience — fast — without giving up the real-time platform that already carries them?

0 Peak concurrent users during the IPL, on a Kafka-backed real-time platform Dream11 already operates.

The bridge

Ship the engagement layer on the platform you already run.

Meridian AI embeds with your team and ships production-grade LLM and agent systems — on your existing stack — in six weeks, not six quarters. For a second-screen experience that means retrieval and function-calling agents wired into your live-match and content services, with a human-review path where it matters, not a parallel system to babysit during the IPL.

One proof

We ship on your existing stack — retrieval over your own data, function-calling agents wired into your current services, and evals plus guardrails in CI. No model training, no rip-and-replace, no new infrastructure to run.

One working session

Map one second-screen workflow, from live signal to shipped feature.

Twenty minutes to pick one interactive moment, define its data, its guardrails, and its human-review path — before we touch code.

See the 20-minute teardown for Dream11