Company profile · 3 min read
Almanac: a shared company brain for humans and agents
Almanac connects company tools into a self-updating wiki that lets agents and teams act from shared context.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
What it does
Almanac is an AI agent and self-updating company wiki. It connects work tools, compiles their context into a shared knowledge layer, and lets an agent act across the same people, projects and decisions. The YC profile describes the product as a productivity suite built for agents; the current homepage frames it as an agent that knows a company rather than a generic chat window.
The fit is a small team with context scattered across email, issue trackers, wikis, CRM and calendars. Almanac is compelling when the cost is repeatedly explaining company history to a new teammate or agent. Its value depends on data freshness, permission boundaries and whether the agent knows when to stop before a login, payment or consequential decision.
Why I’d look closer
The launch description gives concrete workflows: an agent can collect customer feedback into GitHub, prepare a meeting brief, draft follow-up, watch a production environment and raise a pull request. It also says a shared wiki is readable by teammates’ other agents such as Claude Code and Codex. That is a distinct context-layer thesis, but it creates a larger governance surface than a normal notes product.
The founders bring relevant research and systems experience. The YC biographies describe Rohan Sharma with machine-learning research at the Vector Institute, CMU and MIT plus Harvard computational science, Kushagra Chitkara with Harvard data science and IIT Kharagpur, and Divit Sheth with Arcesium engineering experience. The team says it has built agent wikis for Harvard and NASA; that is company-reported background context.
The public pricing page was linked in the source set but did not yield a usable price card in the bounded extraction. That is an unanswered buyer question rather than a claim about what Almanac costs.
What I’d ask
Which sources are indexed, how quickly do changes propagate, and how are permissions inherited when an agent reads shared context? Can a human inspect why an action was proposed, revoke access and export the wiki? I’d start with read-only sources and a low-risk workflow before allowing the agent to edit production systems.
My editorial take
Almanac is a compelling fit for teams whose real bottleneck is context fragmentation. The shared-wiki thesis is more interesting than another assistant. The product earns trust only if the knowledge layer is current, scoped and reversible.
Quick facts
| Field | Sourced detail |
|---|---|
| Buyer fit | Small teams coordinating work across many tools and agents |
| Product | Self-updating company wiki plus agent actions across a productivity suite |
| Integrations | Work tools, email, issue trackers, CRM and calendars; company-described |
| Public pricing | Not established in the fetched pricing surface |
Sources checked
Checked 2026-09-19.
| Source | Used for |
|---|---|
| YC company profile | Product, founders and launch workflows |
| Almanac homepage | Current company-brain positioning |
| Almanac pricing | Pricing surface check |
Cohort context
Almanac 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 2026-09-19T16:18:34.634Z 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 | Not observed in this response |
| H1 or H2 heading | Observed |
| Typed structured data | Observed |
| Docs/developer link | Not observed in this response |
| Pricing link | Observed |
| 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 ↗