I've been talking to a lot of early-stage SaaS teams about how they use AI internally (not the "we're AI-native" marketing version, but the real day-to-day).
What I keep hearing is:
Half the team is on ChatGPT, the other half uses Claude, nobody shares context
Prompts that work well live in someone's Notion doc (maybe)
There's no visibility into what the team is actually using AI for
Onboarding a new hire means "here's a $20/seat subscription, good luck"
We're building something in this space and I'm genuinely curious whether this matches what others are seeing, or if some teams have actually figured it out.
Curious how you deal with gateways going down or getting corrupted after something like an openclaw update? Asking because I ran into this specifically today.
CoChat
@dkofoed we have a monitoring services on the main host watching over the VMs (kind of like a supervisor) but I will admit it needs some tweaking and reconnecting after a dropped connection takes a hot minute. We’re working on improving that.
The “agents as teammates” framing is really interesting.
One thing I’m curious about: when a new team first opens CoChat, what’s the main starting workflow you expect them to try first?
With platforms that combine agents, automation and collaboration, sometimes the hardest part is helping users understand the first simple use case.
Question: Is the "CoChat" name hinting at a real-time collaborative layer over existing chatbots, or is this building something from scratch? The minimalist landing page has me intrigued but also wondering if I'm looking at a wrapper around ChatGPT's API or a completely new conversational architecture.
Origin
Congratulations on your launch, and interesting concept! Caught my eye because recently our org completely banned OpenClaw due to IP/data leakage risks, even if running locally.
I understand that having logs and approvals for sensitive steps is a great way forward, but can you really claim that's all it takes to make it "secure"? Also, what about data retention since agents have memory? Curious how you tackle data retention and leakage risks.
Interesting direction as moving AI agents from isolated chatbots into a shared team workspace feels like a natural next step for collaboration. Curious how teams will balance agent autonomy, memory, and security as these systems start handling real responsibilities.
the memory and personality per agent is the right call. most team AI tools feel stateless, like starting over every session.
curious about the cold start for new team members though. agents build context over time, but what about day one? been using northr identity for that exact gap, portable user identity that loads into any AI session instantly. combined with cochat's memory layer that could be pretty powerful.
Love the idea of mixing AI agents and human conversations in the same thread. That shared context could make AI much more practical for real work instead of just quick prompts. Excited to see where this goes!