Sila Wants AI Agents to Be First-Class Members of Your Team Chat

The Macro: Slack Added AI, but It Still Feels Like a Chatbot in the Corner

Nearly 60% of knowledge workers say they are overwhelmed by their communication tools. The irony is thick. The tools designed to make work faster are making work harder. Slack channels multiply. Messages pile up. Important decisions get buried in threads.

Slack has added AI features. So has Teams. But these feel like bolted-on capabilities rather than native rethinking of how team communication should work. Slack AI can summarize channels and search conversations. Useful, but incremental. It does not fundamentally change the dynamic of humans drowning in messages while AI watches from the sideline.

The question is whether someone can build a messaging platform where AI is not an assistant that you occasionally invoke, but a genuine participant in the workflow. Agents that join channels, participate in discussions, execute tasks, and maintain context across conversations. Not a chatbot you talk to. A coworker you collaborate with.

This is a bold bet. Messaging platforms have massive network effects. Slack has 30 million daily active users. Teams has 320 million monthly active users. Building a new messaging platform from scratch means competing against deeply embedded products that teams use every day.

The Micro: Custom Agents as Native Channel Members

Mith Paresh Patel and Carl Huang founded Sila. Mith is CEO and Carl is CTO. They are a two-person team from YC Winter 2026 working with Diana Hu.

Sila lets you create custom workspace agents with dedicated roles that connect to your apps and workflows. These agents sit in your channels and participate naturally. They are not bots you ping with slash commands. They are always-on participants that talk, listen, and work alongside your team. Collaborative AI sessions let multiple team members work with an agent simultaneously. The agents retain organizational context from conversations and connected apps.

The platform integrates with 99+ services including GitHub, Gmail, Notion, Figma, Linear, HubSpot, Jira, Stripe, and Zoom. You can bulk import from Slack and Teams. Smart search works across all connected applications and channels. There are even external phone numbers for customer communication.

The security positioning is strong. No data used for AI training, SOC 2 certification in progress, end-to-end encryption. The free tier lets teams try the product without commitment.

The Verdict

Sila is making the most ambitious play in this batch. Replacing Slack and Teams is extraordinarily difficult. The switching costs are enormous. The network effects are real. Most messaging startups fail because they underestimate how deeply embedded these tools are in daily workflows.

But Sila has a genuine differentiator. If AI agents as native team members is a compelling enough experience, it could be the forcing function that makes teams willing to switch. The Slack-to-Sila migration is only worth it if the AI capability is dramatically better than what Slack offers natively.

In 30 days, I want to see team adoption patterns. Are entire teams switching, or just individuals trying it out? In 60 days, the question is agent usage. Are the custom agents actually being used daily, or do teams set them up and forget about them? In 90 days, I want to know about the Slack import retention rate. How many teams that import their Slack workspace actually stay on Sila? That number tells you everything about whether the product is genuinely better.