Pulse
Welcome back
Identity-only sign-in. No workspace-level permissions asked — we just need to know it's you.
Sign in with SlackWorkspace hasn't installed Pulse yet? Add to Slack →
Four rituals, one bot: async team check-ins, smart 1:1 intros, channel-aware watercoolers, and a dashboard that tells you whether any of it is landing. AI that learns from what worked — and a one-DM off switch for every person.
Slack-native · Hard opt-outs, self-served · 14-day free trial
Apr 14 – 18, 2026 · 14 teammates replied
Onboarding flow live Tuesday. Three APAC client dashboards rolled out. Timezone-aware scheduler production-tested at 150 people.
Stripe migration waiting on legal (target: early next week). One infra ticket open — owner on leave; backup picks it up tomorrow.
Q3 planning. Two people prototyping an HRIS sync. Genuine interest in an in-person retro surfaced this week.
One click OAuth. No tokens to paste. Pulse syncs your full members list automatically on install.
Start a newsletter, a pairing program, or a watercooler — each has its own schedule, channel, and audience. Nothing is "global" that doesn't need to be.
Pulse scores pair-ups by shared interests, remembers which conversation starters led to meetings, and proposes watercooler questions based on what's alive in your channels. Admins approve before anything is posted.
Four decisions that show up in the copy, the matches, and the trust.
Pulse ranks partners by topic-tag overlap, activity balance, and freshness of engagement. Two shy repliers don't get paired together — and people who share interests actually land together.
After a pair meets, Pulse asks a one-tap "how'd it go?" three days later. Winning conversation starters feed the next round's AI prompt. Every pairing round shows its met-rate vs a rolling 8-round baseline.
When both teammates link Google Calendar or Outlook, Pulse finds a mutually-free 30-minute window in the next 7 days and drops a Zoom link right into the intro DM. Their consent, every time.
Pulse reads your chosen source channel daily, extracts themes, and proposes questions tailored to the week's vibe. Admins approve in a queue before anything posts — with permalinks to the inspiring messages.
Turn each on when you're ready. Every ritual respects the same hard opt-outs, timezone-aware send windows, and self-service DM commands.
Each question gets its own DM thread per teammate. AI writes a personalized greeting using what it knows about them; follow-ups are individually crafted, not templated. A consolidated digest publishes to a Slack channel you choose.
Create programs tied to a channel and audience. Pairs or triads, your call. AI writes a "why this pair" brief and a starter tailored to their shared topics. Post-meeting feedback feeds the next round.
Each watercooler is bound to a channel. The bot posts the question, then drops its own first reply 60 seconds later to break the ice. PACK badges on sent prompts show what's landing.
Participation rate, meeting rate, watercooler response rate, AI suggestion acceptance — stoplight colors. Plus dormant-users list with a one-click "send a thoughtful nudge" action that references their known interests.
How does opting out work?
Any teammate DMs the bot pause / resume, snooze 2 weeks, don't pair me with @person, or delete my profile — no admin needed. Hard gates: if your flag is off, nothing gets sent, regardless of what the AI "wanted" to do.
How does the AI decide who to pair?
A small, explainable scorer: 55% shared topic tags, 25% complementary activity levels (mix a heavy replier with a medium one), 20% recency of engagement. Hard constraints (no-repeat window, private avoid list, snoozes) are enforced before scoring.
How transparent is the AI layer?
Every AI call writes a row to an ai_traces audit table — agent, model, tokens, cost, latency. Admins can see exactly what the AI is doing and what it costs. Providers are swappable and features degrade gracefully if no AI key is set.
Where does data live?
Your workspace's Pulse instance holds its own database. Slack tokens, OAuth tokens for calendar integrations, and any secrets are encrypted at rest. Nothing is used to train external models.
Does the bot ever post without admin approval?
Bot-drafted watercooler questions go into a review queue — an admin approves (with optional edits) or rejects each one. Personalized DMs go out only within rate limits (1 DM per user per 7 days, 1 thread follow-up per round per user).
What topics won't the bot touch?
Health, family, finances, religion, politics, performance reviews, and trending crises are hard-coded blocks in every AI agent's instructions. It paraphrases past replies rather than quoting them.
Can I bring my own OAuth apps?
For calendar and Zoom integrations: each workspace can register its own OAuth apps with those providers, so the consent screens show your company name, not Pulse's. Configured in the admin Integrations page.
Where does timezone intelligence come from?
Slack's users.list returns each member's timezone. Pulse stages pairing DMs to land inside each recipient's preferred send-hour window — one person doesn't get 2 AM pings because the other's in a different tz.
14-day free trial. No credit card. Install from Slack, turn on the rituals you want, keep the ones your team likes.
Donut is a solid product. Here's where Pulse is different — and where we're not.
Comparison reflects Pulse features that are live in the codebase today. Donut rows reflect their published documentation at the time of writing. We'll update this as either product changes.