Company profile · 3 min read
Assemble: reversible, auditable agent work inside enterprise systems
Autonomous IT delivery for ERP, CRM, HRIS, and other core systems with dependency mapping, validation, review, and rollback.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
Assemble is building autonomous IT delivery for teams changing ERP, CRM, HRIS, and other core business systems. The buyer decision is whether an agent can scope dependencies, implement a change, validate it, and leave a reversible trace without bypassing the review and rollback controls enterprise IT needs.
What it does
Assemble says its agents map how systems operate, implement changes inside enterprise platforms, verify the result, and record every action. The homepage demo shows a Salesforce-to-Oracle route change with a review gate, validation, rollback, and trace. The public launch names Salesforce, SAP, Workday, NetSuite, and custom applications as target systems (Assemble homepage; YC launch).
| Fact | What the public sources say |
|---|---|
| Buyer | Enterprise IT, business-systems, applications, and transformation teams |
| Workflow | Scope dependencies, implement changes, test, audit, review, and rollback |
| Systems | ERP, CRM, HRIS, Salesforce, SAP, Workday, NetSuite, ServiceNow, and others are named |
| Control model | Human review, sandbox validation, rollback, and agent traces |
| Founders | Aliyan Ishfaq, Shaurnav Ghosh, and Shrish Janarthanan |
Why it fits
The product targets the gap between a ticket and a safe production change. Enterprise systems are full of hidden dependencies; changing one route or field can break downstream workflows. Assemble's promise is to make the dependency map, change set, test evidence, and rollback path visible before a human approves promotion.
The homepage claims up to 40× faster work and half the IT spend, but those are company claims rather than an independent implementation benchmark (Assemble homepage). A buyer should test a non-production change across its real integration graph, permissions, test data, rollback behavior, and CAB evidence. Pricing was not published, and “100% verified” should be understood as a company product claim, not proof that every enterprise change is safe.
The founder background fits the problem: YC describes Ishfaq and Ghosh as Stanford AI-lab researchers with LangChain, AWS, and Apple experience, while Janarthanan brings cybersecurity and IT-team context (YC company profile).
Short version: Assemble is worth a controlled sandbox evaluation for a team with an enterprise-app backlog. Start with reversible changes and keep production approval human-owned.
Sources checked — 2026-09-19
Cohort context
Assemble is listed in Summer 2026. In our 2026-09-18 directory snapshot, 119 of 232 listed companies in that cohort have YC’s primary industry label B2B (51.3%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.
Public website snapshot
Observed 2026-09-19T16:18:35.043Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.
| Signal | Homepage observation |
|---|---|
| Product description metadata | Observed |
| Canonical link | Observed |
| H1 or H2 heading | Observed |
| Typed structured data | Observed |
| Docs/developer link | Not observed in this response |
| Pricing link | Not observed in this response |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |
Public observations · Collection method. Missing links here do not establish that a capability or file is absent elsewhere.
About the author
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.
First1000 ↗ · X ↗