Two tools can look nearly identical in a spec and carry completely different risks. One drafts marketing emails. One scores loan applications. The first can be edited before it goes out; the second can affect someone's access to credit.
1
2
118
So the first governance questions belong at decide, not at a final approval meeting: who owns this, who could be hurt if it fails, and what should require a person?
1
56
Then put those answers in the spec. The people who understand the rule can define it before an agent builds from it. Checks during build and a version-specific check at ship become ways to enforce that decision, not substitutes for making it.
Read the post: sycamore.so/blog/company/gov…
16
A sign-off is not a transferable sticker. It belongs to one version of a change. Edit the change after approval and the old approval should not silently follow it.
1
2
92
At ship, the question can be small and concrete: is this exactly the version that was approved? Check what is actually deployed, not only what the release plan says.
1
44
The work does not end at ship. At run, someone has to watch what the system and its agents do and be able to stop them. At improve, compare what happened with the spec and feed the lesson into the next decision. That is governance across the work, rather than a single queue at the end.
Read the post: sycamore.so/blog/company/gov…
16
Our Founder and CEO @SriViswan joined Tobi to talk about building an operating system for enterprise AI agents, and why he believes the more governance you have, the faster you can go.
Listen now: alphalist.com/podcast/149-14…
The first version is a beginning. The fiftieth change reveals whether the software can keep up. Each policy and exception needs people who know the system, plus testing, approval and release. That cost turns good ideas into a backlog.
1
1
203
If changes get cheaper, teams can make smaller revisions. Imagine a bank checking a second-review rule for queues and revising its threshold while the decision is fresh.
1
73
Each change should carry forward connections, approvals and tests so the next one is easier to make and review. The fiftieth should cost less than the first. We intend to measure whether that curve bends.
sycamore.so/blog/company/ent…
29
A policy changes in an hour. The software might take a quarter to follow. In that gap, people make a spreadsheet, share an inbox and keep doing their jobs. Eventually the official system records decisions that happened elsewhere.
1
1
174
Imagine a bank adding a second risk review above a loan threshold. While the form and approval path wait for changes, the team tracks loans in a sheet and emails risk. Another rule adds a tab.
1
84
The workaround makes sense for a day. After months, it is hard to see who approved what or which rule applies. Enterprise software should absorb small, reviewed changes while the business decision is still fresh.
sycamore.so/blog/company/ent…
39
Sycamore Build is now generally available. We've been working closely with enterprises to build software that makes their businesses AI-native. Today, we’re excited to bring that capability to everyone.
We believe the people who understand a business best should be able to build the software that runs it. With Sycamore, they can.
Describe what you need, provide context, and collaborate with Build to create, test, and deploy enterprise-grade agents and applications, purpose-built for your organization.
And when your processes change, your software changes with them. Every app has self-improvement built in, so it keeps getting better over time. It's the end of static, one-size-fits-all enterprise software.
Our vision is much bigger: a platform where agents run enterprises autonomously, with humans in control, and software that continuously improves and earns your trust. Sycamore Build is our first step toward that future.
We're incredibly excited to see what you build!
Get started at sycamore.so. The first 1,000 signups using code LAUNCH500 get $500 in credits!
Sycamore retweeted
We're hiring across 10+ roles in Palo Alto in Applied AI Engineering, Forward Deployed Domain Experts, Product Engineering, Data Systems, Infrastructure, Cloud, Security, and Research. Apply now!
Sycamore is proud to sign the Open Weights and American AI Leadership letter.
We build our platform so any team can turn intent into enterprise-grade agents. The intelligence behind that software should be portable, adaptable, and owned by the teams building on it. Open weights make that possible.
Since launching last week, more than 230 companies and organizations from across the tech sector have signed the "Open Weights and American AI Leadership" open letter. We want to thank these partners for standing up and publicly supporting broader access to AI innovation. A special thanks to @nvidia, @a16z, and @PalantirTech for working with @Microsoft on this effort.
These signatories understand that America’s AI leadership will not depend on the success of our frontier models alone, but on our ability to build a strong, secure, and open ecosystem that diffuses AI into every sector. We look forward to continuing to work with our partners and with policymakers to build that open ecosystem in a way that benefits American businesses, empowers American workers, and strengthens the American economy.
Sycamore retweeted
Listen to my conversation with @SriViswan of @SycamoreLabs
Spotify: bit.ly/3v1R0Tu
Apple: bit.ly/4bTCwpD
Youtube: bit.ly/3uXthnv
LinkedIn: bit.ly/3Xs8GQP
#EnterpriseAI #Startups #VC #CloudScale #AIAgents
Sycamore retweeted
“No man steps into the same river twice. It is not the same river, and it is not the same man,” @SycamoreLabs founder @SriViswan shares. “It also applies to Sycamore. No Sycamore software runs twice in the same enterprise, for it is not the same software.”
Enterprise software is outdated. Sycamore is building AI-powered systems to make companies autonomous using a dynamic agentic AI operating system that constantly improves itself.
“Every time our agent runs in an enterprise, it updates itself and gets better.”
From The Investment Memo, with @ravirajjain.
Some of the best founders take unconventional career journeys.
@SriViswan is no exception. His diverse and extensive career experience led him to build @SycamoreLabs, an operating system for enterprise AI agents.
We recently co-led his $65M seed raise alongside @coatuemgmt and had the pleasure of talking with him on our latest episode of The Investment Memo. Sri and Lightspeed Partner @ravirajjain discussed the current market gap for enterprise AI software and how Sycamore differentiates. Check out the video below to hear more.
0:00 Introduction to The Investment Memo
3:44 Reflections on Enterprise Software Challenges
5:50 The Current Enterprise Software Landscape
8:07 Enterprise Requirements: Security, Trust, and Compliance
12:15 Sycamore’s Agent Operating System and How Software Improves Itself
14:07 The Market Opportunity for Sycamore Labs
17:02 Building at Machine Speed: Sycamore's Development Process
20:52 From Two-Week Sprints to Daily Cycles
25:52 Why Lightspeed as an Investment Partner
27:29 $65M Raise and Future Milestones
28:47 Closing Remarks
Sycamore retweeted
"In the morning we decide what needs to be built. Every person has up to five or seven agents working with them. By evening we set it all up so that when we're sleeping, the agents do the work."
@SriViswan (Founder and CEO of Sycamore Labs), on what building software at "machine speed" looks like:
"Our productivity is measured by how we can unlock agents. Every day we come in and say: what can we do today so we can unleash the agents to do better, faster, cheaper?"
"We're a 24/7 company. Not for humans, but for agents."
From The Investment Memo with @ravirajjain
Sycamore retweeted
I'm thrilled to announce @SycamoreLabs, the trusted agent OS for the enterprise. We raised a $65M seed led by @coatuemgmt and @lightspeedvp, with @AbstractVC, @DellTechCapital, @8vc, @Fellows_Fund, @e14fund, and an exceptional group of angel investors.










