Product
Everything the job needs, in one space
Threads, docs, and tables. The tools already connected. Your team and Hops, working from the same memory.
What's inside
One space holds the whole job
Conversations, shared memory, structured data, and the handoffs between the team and Hops. Not a chat window bolted onto your tools: the place the work actually happens.
Spaces
A space for every team and project
Threads, docs, tables, files, and the connected tools for one team or one project, in one place. Invite people, invite Hops, and the permissions of the space apply to both.
- Chat, DMs, and threads your team already knows how to use
- Hops only works where it's invited
- Pins and links keep the important things one click away
Shared memory
Memory the whole team can see
Decisions, changes, and conversations stay connected to the work. Hops reads the same docs and tables your team edits, and what it remembers is visible and editable, not locked in a private chat.
- Docs and tables both kinds of coworkers edit
- Every draft shows its sources
- Inspect, edit, or remove what's kept
q2-launchLaunch Readiness Memo
EditingShare
Launch Readiness Memo
Where we are against the Q2 launch plan, what's blocking, and who owns the last mile. Hops keeps the numbers current from the checklist table.
Ship agent-native chat to GA
- Public launch by Apr 18 — on track
- ≥ 30% weekly active use of the agent surface within 4 weeks
- P0 incident rate ≤ 1/month through launch week
Tighten the pricing story
- v4 pricing block live before announcement
- Enterprise floor held at 50-seat minimum
- File-search add-on decision shipped
Tables and apps
Structured data, and the views built on it
Orders, pipelines, contracts, and projects live next to the conversations about them. Hops queries and updates the rows, and builds the board, dashboard, or one-pager you ask for on top, inside your permissions.
- Tables Hops keeps honest
- Dashboards and boards built on request
- Nothing leaves the space's access
Launch dashboard
Built by Hops · from Q2 Launch ChecklistSignups this week
1,284+18%
Activation rate
46%+3pts
Open P0 incidents
0
Signups per week
Handoffs
Handoffs that carry their context
Work moves between people and Hops without a re-brief. A teammate, or Hops, picks up the thread with the sources, owners, and next steps already attached.
- Work returns to the thread, sources attached
- Pass it on without re-explaining
- Nothing sends until a person approves
q2-launch
31ChatDocs & TablesAutomations
Today
9:02 AM
Morning — @Hops where are we against the launch checklist?
AGENT9:02 AM
Two of eight tasks are done; one is blocked. Summary below — details are in the table.
Q2 launch — status
From the launch checklist · just now
- Done 2 of 8 tasks
- Blocked Launch readiness sign-off (comms freeze)
- Next v4 pricing block draft — due Apr 27
The agent
Not a sidebar. A teammate in the thread.
@Hops is built in, in every space it's invited to. Install skills for the specific jobs, and share the ones that work with your team. It answers where you ask, reads what the space already knows, and shows its work.
atlas-renewal
Nina Cole
@Hops where does the Atlas renewal stand?
Hops
At risk. Usage down 22% since March, and the champion left in May.
Hops
Q3 forecast adjusted to match.
In the conversation
Mention @Hops the way you'd mention a teammate. It answers in the thread, picks up the context around the request, and hands the result back where everyone can see it.
Docs & Tables
NameAuthor
Renewal tracker
Q3 pricing deck
Atlas call notes
Working from shared context
The docs, trackers, and memory your team edits are the same ones Hops reads and keeps current. No briefing doc, no pasting context around.
MessagesFree
TeammatesFree
HistoryFree
Agent work$49
Metered where it matters
Messages, teammates, and history are free. You pay for agent work, not for talking about it.
Hops at work
Routine work on a schedule. Sensitive work on hold.
Hops does the recurring jobs and stops before anything that matters. Every run is visible, and every approval is a person's.
Automations in plain language
Describe the rule the way you'd say it out loud: every Monday, when a deal stalls, after a PR merges. Hops runs it on a schedule or a trigger and posts the result where the team can see it.
Automations
Pipeline reviewRunning
Renewal prepAug 14
Month-end closeJun 30
Approvals before anything sensitive
Sending an email, changing a stage, paying a bill: Hops drafts it and waits. The action stops in the thread until a named person approves, edits, or declines.
Send renewal quote
Atlas Corp · $48,000
Needs approval
Waiting on Maya LindbergSandboxed execution
Hops runs real software (ffmpeg, LibreOffice, full toolchains) in isolated sandboxes. It's how a message comes back as a file.
Your data, scoped
Hops sees what the space sees, nothing more. Access follows the same permissions your people have.
Audit log
Every action Hops takes, the sources it used, and who approved it are recorded, so the work stays inspectable after the fact.
Centralized AI spend
Every skill, model call, and permission in one place. One bill instead of a drawer of AI subscriptions.
2000+ Integrations
Plugs into the tools your team already runs
Where this is going
AI started as a private chat: one person, one assistant, one context window. Team work doesn't fit in there. Hops is built the other way around: your team and Hops in the same place, working from the same memory, with the work in the open.
Most AI tools start with one person and one chat. Hops starts with the team. Hops reads the threads, finds the docs, remembers the decisions, and hands off work with the sources attached. People stay in control of every send. That's the product today, and it's the direction everything we build follows.
FAQ
Questions, answered
What is Hops?
Hops is the platform where people and one agent get work done together. Your team and Hops share one place, one memory, and one set of permissions, instead of everyone re-briefing a private chatbot.
How is it different from a chat assistant?
Most AI tools start with one person and one chat. Hops starts with the team. Hops works inside the spaces your team already uses, reads the context that's there, and posts its work where everyone can see it.
Is there just one agent?
Yes: @Hops, built in, working in any space it's invited to. Skills teach it specific jobs (sales, finance, research). Install them for yourself and share the ones that work with your team.
What can Hops actually do?
Draft, flag, prep, brief, pull, match, sync, post, and redline. It reads threads and docs, queries tables, pulls from connected tools, and returns the work to the thread with sources. Anything sensitive waits for a person.
What does Hops remember?
Decisions, changes, and conversations connected to the work in its space. Memory is visible to the team, and you can inspect, edit, or remove what's kept. It doesn't reach into spaces it hasn't been invited to.
Does Hops act on its own?
Only on the routines you set up, and never on a sensitive action. Automations run on a schedule or a trigger, and anything like sending, paying, or changing a record stops as a draft for approval.
Does it connect to our tools?
Yes. 2,000+ integrations, added from the composer and used through the permissions the space already has. Hops reaches the tools your team runs, not a copy of them.
Who can see what?
Permissions live on the space. People and Hops in a space see its threads, docs, and tables, and Hops sees nothing outside it. Admins set policies once and they apply everywhere Hops works.
One AI teammate for the whole team
Knows the work. Forgets nothing. Shares with everyone.