Decimal Life is now part of AWS Activate and Sarvam AI for Startups

Which startup programs is Decimal Life part of?

Two. Both are programs for early-stage companies, and both belong to companies whose technology Deci already runs on.

Why AWS and Sarvam AI?

Neither program is a logo we picked off a list. Both are the companies Deci already depends on, every minute it runs.

Amazon Web Services is where Deci lives. Every time you open Deci, the request lands on AWS in the Asia Pacific (Mumbai) region. The API, the workers that keep your Life graph current, the Postgres database that holds it and the cache that keeps it fast all run there. AWS Activate is Amazon’s program for early-stage startups, offering credits, technical support and training to teams building on AWS.

Sarvam AI is how Deci thinks. Sarvam is an Indian AI company building large language models, and Deci’s conversations run on Sarvam’s 105B models — the everyday chat model and a deeper reasoning one for harder questions. When you ask Deci what your day looks like, a Sarvam model is reading the context your Life graph assembled and writing the answer. Sarvam AI for Startups is its program for early-stage teams building on those models.

So joining both programs doesn’t change our stack. It supports the one we already chose.

Why did we build Deci on this stack?

Deci is a personal context engine: it turns your calendar, inbox and health into one Life Knowledge Graph, so AI can start from your actual day instead of a blank chat. That puts two demands on the stack, and they pull in opposite directions.

It has to be boringly reliable, because a planner that is down at 8 a.m. is worse than no planner. And the AI has to be good at the messy, mixed-language way people actually write — the email that switches between English and Hindi halfway through, the task called “call mummy re: Diwali tickets”.

AWS answers the first. Sarvam, building models in India with Indian languages in mind, answers the second. We also keep a second AI provider standing by, so Deci keeps answering even if a model is ever unavailable.

What does it mean for people using Deci?

Mostly, you won’t notice — and that is the point. What changes is on our side:

• More of the budget goes into Deci. Every rupee not spent on servers and model tokens goes into the planner, Inbox Radar and Health OS.
• Room to make Deci think harder. Better context means longer prompts and more careful reasoning. The programs give us headroom to do that without cutting corners.
• Fewer trade-offs on reliability. Backups, monitoring and spare capacity are the first things a small team trims to save money. We would rather keep them.

What doesn’t it mean?

We want to be precise here, because trust is the whole product.

What’s next for Deci?

Deci is in private beta, and we are letting people in steadily. With these two programs behind the stack, the next few months are about depth: a planner that understands your week better, an inbox that knows which threads actually matter, and a Life graph that keeps getting more useful the longer you use it.

Thank you to the AWS Activate and Sarvam AI for Startups teams for having us. And thank you to everyone already using Deci — you are the reason any of this is worth building.