Hey everyone, another question from the Velozity team as we get closer to launch.
Last time we asked how AI work moves between people on a team. This time we want to dig into trust.
Right now most people keep their AI on a short leash: it drafts, it suggests, but a person clicks send. That's safe, but it also means every handoff between a person and their AI, or between two people's AIs, still needs a human in the loop to move things along.
We're building Velozity so agents can work more directly inside a team's shared space, chat, tasks, files, calendar, with a person approving anything that leaves the workspace rather than approving every single step along the way.
Solving such a common problem that people hardly even thinks as a problem but it eases the hassle and optimize the work soo much that we actually realize how big of a problem it was. The execution is phenomenal. Kudos !
@pratham_aggarwal2 Thank you, Pratham. That is the part we find hardest to explain: nobody calls it a problem until they see their team and their agents working in one place, and then it is hard to go back. Really glad the execution stood out. If anything feels rough as you use it, tell us, we are fixing things daily.
@pawani_krishna Thanks, Pawani! Yes, it works well for study groups. Your group's chats, notes, and AI help all live in one shared space instead of being scattered across apps. There's a free plan to get started, so you can try it with your group at no cost.
@pawani_krishna
A good way to start is one Space per subject or study group. Drop in your notes and lecture material, and everyone in the group, plus any AI agent you add, works from the same set. If you record a group call there, the discussion stays in the Space too, so nobody has to rewrite what was covered. Would love to hear how it goes for your group.
Love the bring your own AI part. My team already pays for our own tools, and I really don't want another bill. Quick question though. I run an SEO agency with a bunch of clients. How do you keep one client's stuff from leaking into another client's agent work?
@michaelgleed Thanks, Michael, and good question for an agency. The way to set it up is one space per client. An agent only sees the conversation, files, notes, call transcripts and tasks of the space it's been added to. So if your Client A agent isn't added to Client B's space, it has no way to see Client B's work. Each client's work stays in its own room. And since agents run on the AI subscriptions your team already pays for, there's no extra bill. How are you handling this for your clients today?
@michaelgleed
The same rule covers the people side too. If a client joins their own space, they see only that room, so you can work with them directly without exposing your other accounts. For an agency, a simple habit helps: name each space after the client, so it is always obvious which room you and the agent are working in.
Saw that WhatsApp Business is one of the connectors. Can an agent draft a reply to a customer there and send it through the same approval step, or is it read-only context for now? That's where a wrong auto-reply would hurt the most.
@lancelot_d_souza1 Good question, Lancelot, and you have picked the right place to be careful. Today WhatsApp Business is a send connector, not read-only context. An agent can draft a reply and send it from your business number, and it shows you the recipient and exact text and waits for your go-ahead first. Nothing goes out as an auto-reply. The limit is on the other side: an agent can't read incoming customer messages on WhatsApp yet, so you bring the customer's message into the conversation and the agent drafts from there. Reading inbound messages is the piece we still need to add.
@ishudarshan That's the right order to build it in, honestly. Send with approval first. When inbound reading lands, will the agent see the full conversation history or only messages from the point it's connected?
Pulling chat, calls, tasks, calendar and files into one workspace and then letting your existing AI subscriptions work inside it is a sensible direction, since the problem is usually fragmentation rather than missing features. Sharing the same context across agents is the detail that makes it useful. Do the agents keep awareness of meetings and tasks across the whole space?
@karimbenkeroum Thanks, and you've put it well: fragmentation is the real problem. Yes, they do.
An agent in a space can see that space's conversations, files, notes, call transcripts shared there, and tasks. So if something was agreed in Monday's meeting, the agent knows it when you hand it a task on Tuesday. That awareness is scoped to the space the agent is in, so it doesn't spill into spaces it wasn't added to.
@karimbenkeroum
It stays current too. An agent can check what changed in the Space recently (new files, shared call transcripts, note edits, task updates), so it picks up where the team is today and not from a stale snapshot. Your own calendar is separate: an agent reaches it only through your access, which is how it can prep you before a meeting without exposing your schedule to the rest of the Space.
"Interesting take. Most "AI teammate" tools I've seen are just a chat box bolted onto a workspace. Having the agent sit in the same space as the team, where everyone can see what it's doing, is a different idea. What happens when two people ask the same agent for conflicting things at the same time?"
@suryansh_tiwari2
When two people ask for conflicting things, the agent doesn't pick a winner on its own. It flags the conflict in the shared space, shows both requests, and waits for someone to decide. Nothing goes out without approval, so a clash never turns into a wrong action.
@suryansh_tiwari2
One more thing that keeps this rare, Suryansh: every agent has one owner, and only that owner can give it work. So two teammates pulling the same agent in opposite directions doesn't really happen; each person works through their own agent in the shared Space. When two agents do land on the same piece of work, each can see who else is working there, and a second agent is stopped from overwriting a file another one is editing.
Congrats, Ishu and Akash! Keeping the agents in the same context as the team makes a lot of sense. Re-explaining the work every time you switch tools is the part I'd want to lose.
@oleksii_sekundant Thank you, Oleksii! 🙌 That's exactly the thing we wanted to fix. Most of the time lost with AI isn't the work; it's explaining it again every time you switch tools. In Velozity, the agents sit in the same room as the team, so they pick up where the team left off. Which tool switch costs you the most today?
Thanks@oleksii_sekundant
Glad that part landed. Adding to Ishu's note: the re-explaining goes away for people too, since a teammate who joins late reads the same room the agent works from. One place to catch up, for everyone.