Your team already has too many tabs open. Adding another app to the stack is easy to justify in a demo and hard to sustain in practice. The tools that actually get used are the ones that show up where people already work.
For most teams, that place is Slack. Here is why that matters for choosing an AI assistant:
Every tab switch is a decision not to use the tool
This sounds minor but it adds up. When someone has a question and the AI tool is in another tab, they have to stop what they are doing, switch context, type the question, get the answer, switch back, and paste it into the conversation where it was needed.
Most people will not do that for a quick question. They will just ask a colleague instead, or skip the question entirely. The tool sits unused. Not because it is bad, but because it is inconvenient.
When the tool is in Slack, asking a question is the same as typing a message. There is no tab to switch to. There is no login to remember. The answer appears in the same conversation where the question came up. This is what that looks like:

@Palfred where's the onboarding checklist? Starting week two and want to make sure I haven't missed anything

Found it in Confluence — here's your checklist:
✓ Dev environment setup (Day 1)
✓ Security training (Day 2)
✓ Meet your onboarding buddy (Week 1)
○ Shadow a team lead (Week 2)
○ Complete first project (Week 2)
You're right on track. I've taken the liberty of checking — your team lead has Thursday afternoon open for a shadow session. Would you like me to book it?

@Palfred send the product update email to the Q1 launch segment

Ready to send. I've pulled the segment from HubSpot — just need your go-ahead:

Good morning. I've noticed something worth your attention on the Acme Corp account:
I've drafted a check-in email for their account manager. Would you like to review it before I send?

good catch — send it. Also loop in @james before the renewal call

@Palfred what are users asking for most right now?

I've reviewed Intercom conversations and matched them against your Linear roadmap:
1. CSV export for reports · 34 mentions
● On roadmap — Sprint 14
2. Slack notifications for form submissions · 21 mentions
⚠ Not on roadmap
3. Team-level permissions · 18 mentions
● On roadmap — Q2
If it helps, I can put together a one-pager for the next prioritization meeting.

@Palfred we're getting a spike in tickets about the export feature — what's going on?

Checked Zendesk — 12 tickets in the last 6 hours about CSV exports timing out.
I can have a Jira issue raised and the on-call engineer notified in a moment.
Answers in threads, not in silos
When someone gets an answer from a separate AI app, that answer lives in their personal chat history. Nobody else on the team sees it. If someone else has the same question tomorrow, the cycle repeats.
When the answer shows up in a Slack thread, it is visible to everyone in the channel. It becomes searchable. It is part of the team's shared context. The next person with the same question might find it without even asking.
This changes how knowledge moves through a team. Instead of answers disappearing into individual chats, they accumulate in channels where the work actually happens.
Adoption is not a training problem
When a new tool requires onboarding sessions, training docs, and Slack reminders to “please use the tool,” adoption is already failing. The tool is fighting against people's habits instead of fitting into them.
The teams that see the fastest adoption are the ones where someone mentions @Palfred in a channel, gets a useful answer, and a colleague sees it happen. No training required. The tool demonstrates its own value because it works in the place where people are already looking.
Your integrations should meet you where you are
Pulling data from GitHub, Jira, Salesforce, or Notion is useful in any tool. But the value increases when that data lands directly in the conversation where a decision is being made.
A sales rep asking about a deal in Slack gets the Salesforce data right there, in the thread where the account team is discussing next steps. An engineer checking a PR status gets the answer in the channel where the team is coordinating the release.
The data does not just get retrieved. It gets used, immediately, by the people who need it.
If your team lives in Slack, your AI assistant should too. Not as a notification bot that points people somewhere else, but as a tool that works natively in channels, threads, and DMs. That is why we built Palfred the way we did. Not as a standalone product with a Slack integration, but as a Slack app from the ground up.