Sutra reasons across your ERP, PLM, MES, Slack, and Email, answering engineering questions in seconds, simulating the downstream impact of every change, and executing the follow-on work automatically.
5 years ago, me and @hemanthug didn't know we will be building a company together.
It started with a script to fill excel sheets at Hemant's hardware job, which turned into @Sutra
we have been at it for the past 2 months
Discovery conversations with engineers from Figure AI, Rivian, Caterpillar, Reliable Robotics and Nova Semiconductors told us the problem is much bigger than we imagined.
3 LOIs, 1 live, 2 onboarding. and a check from an a16z scout, we are just getting started.
Hardware manufacturing is having its moment right now, sutra lets hardware manufacturing teams make decisions faster and make them better. We are bringing the infrastructure that accelerated software development with AI to the manufacturing industry.
I have lived this exact problem for the past 3 years. As someone with a MS in Aerospace from UIUC I spent 90% of my engineering time filling spreadsheets and sending emails. I looked for a solutions but nothing existed. So we built it.
Sutra is what every engineer we talked to wish existed, it sits above a hardware team's systems of record (PLM, ERP, MES, email, spreadsheets) and answers operational questions, simulates the downstream impact of changes, and executes the follow-on workflows. The architecture mirrors the way engineering decisions actually get made: question, consequence, action.
Today we're launching our agentic layer, which as as an AI PLM analyst and runs your workflows autonomously. Head over to http://heysutra.com/agents to grab a spot on our list, as we are rolling deployments out over the week. We will be hanging out in the comments today!
@thamibenjelloun Thanks! Cross-functional visibility is the point, so it's less about preventing leakage and more about role-scoped access. The underlying graph traversal happens across all connected systems, but what surfaces to a given user is filtered by their role. A qual engineer sees re-qualification triggers, procurement sees inventory and supplier exposure — same change event, different slice.
When a change is proposed, how do you actually compute “downstream impact” (affected BOMs, open orders/WIP, inventory exposure, re-qualification steps, schedule shifts)—and what do you do when the underlying data across PLM/ERP/MES disagrees or is stale?
@curiouskitty BOM traversal, cost impact, schedule risk, and qualification check run in parallel across PLM, ERP, and MES, normalized into a canonical schema so cross-system reasoning is consistent.
On conflicts — each field type has a designated system of record, example, rev state defers to PLM, inventory to ERP, WIP to MES. Where source of record is ambiguous, impact is computed against the conservative value. Every decision is logged against the data that produced it.
Replies
Sutra
5 years ago, me and @hemanthug didn't know we will be building a company together.
It started with a script to fill excel sheets at Hemant's hardware job, which turned into @Sutra
we have been at it for the past 2 months
Discovery conversations with engineers from Figure AI, Rivian, Caterpillar, Reliable Robotics and Nova Semiconductors told us the problem is much bigger than we imagined.
3 LOIs, 1 live, 2 onboarding. and a check from an a16z scout, we are just getting started.
Sutra
I'm Hemant, Co-Founder & CEO of @Sutra.
Hardware manufacturing is having its moment right now, sutra lets hardware manufacturing teams make decisions faster and make them better. We are bringing the infrastructure that accelerated software development with AI to the manufacturing industry.
I have lived this exact problem for the past 3 years. As someone with a MS in Aerospace from UIUC I spent 90% of my engineering time filling spreadsheets and sending emails. I looked for a solutions but nothing existed. So we built it.
Sutra is what every engineer we talked to wish existed, it sits above a hardware team's systems of record (PLM, ERP, MES, email, spreadsheets) and answers operational questions, simulates the downstream impact of changes, and executes the follow-on workflows. The architecture mirrors the way engineering decisions actually get made: question, consequence, action.
Today we're launching our agentic layer, which as as an AI PLM analyst and runs your workflows autonomously. Head over to http://heysutra.com/agents to grab a spot on our list, as we are rolling deployments out over the week. We will be hanging out in the comments today!
Mailwarm
Congrats on the launch. How do you handle permissions so it doesn’t leak info across teams?
Sutra
@thamibenjelloun Thanks! Cross-functional visibility is the point, so it's less about preventing leakage and more about role-scoped access. The underlying graph traversal happens across all connected systems, but what surfaces to a given user is filtered by their role. A qual engineer sees re-qualification triggers, procurement sees inventory and supplier exposure — same change event, different slice.
Product Hunt
Sutra
@curiouskitty BOM traversal, cost impact, schedule risk, and qualification check run in parallel across PLM, ERP, and MES, normalized into a canonical schema so cross-system reasoning is consistent.
On conflicts — each field type has a designated system of record, example, rev state defers to PLM, inventory to ERP, WIP to MES. Where source of record is ambiguous, impact is computed against the conservative value. Every decision is logged against the data that produced it.