Every time you jump from Slack to Jira to Stripe and back again, you're hemorrhaging time. That's not hyperbole—a 2024 study commissioned by Salesforce and Slack pegged the daily cost of context switching at 96 minutes per worker. Nearly two hours, gone. Not to deep work, not even to email. Just... toggling.
Bubble Lab, a freshly minted Y Combinator company run by two Cornell Tech graduates, thinks the fix isn't another tab. It's eliminating the need to open one in the first place.
Their product is called Pearl, and it's an AI assistant that lives entirely inside Slack. The idea: you prompt it once—"Pull all open invoices from Stripe and send me a summary"—and it executes across Notion, Jira, HubSpot, GitHub, and roughly two dozen other connected tools. If the task proves useful, you promote it to an autonomous workflow that runs on its own, no further input required.
According to Bubble Lab's homepage, more than 4,000 business professionals are using Pearl. The company claims setup takes under a minute—a direct OAuth install, not the laborious onboarding rituals that plague enterprise software. Whether those numbers hold up under scrutiny is another question, but the pitch itself is sharp: Slack as command center, Pearl as the engine that makes it work.
The Pivot from Nutrition Tracking to Workflow Automation
Cofounders Selina Li and Zach Zhong didn't start out building ops tools. Their first swing was gymii.ai, a social-first nutrition tracking app that launched in April 2025. By early 2026, they'd abandoned it entirely.
The pivot wasn't a panic move. Li posted on social media that they'd found early traction with an open-source, TypeScript-first approach to automation before applying to YC. That became the foundation for Pearl. Both founders hold degrees from Cornell Tech; Li also studied at UPenn. They went through Y Combinator's Winter 2026 batch, securing the accelerator's standard deal: $500,000 in funding, structured as $125,000 for 7% equity via a post-money SAFE, plus $375,000 through an uncapped MFN SAFE.
Demo Day reportedly took place on March 24, 2026. YC's official LinkedIn account highlighted the launch around late February or early March, framing Pearl as a tool that "supercharges your ops work in Slack." For a two-person team stepping into a crowded market, the timing matters. They're fundraising now, and the clock on YC momentum runs fast.
What Pearl Actually Does—and What It Promises
Pearl's interface is deceptively simple. You talk to it in natural language, inside a Slack channel. "Draft an email to the finance team about Q1 budgets." "Create a Jira ticket for the bug Sarah reported." "Sync this week's sales data from HubSpot into Google Sheets." Pearl handles it, pulling from your connected tools without forcing you to leave Slack.
The real trick comes when you start repeating tasks. That's when Pearl offers to promote the action into a scheduled, autonomous workflow. Bubble Lab's site claims these promoted workflows run "~10x faster," cost "~90% cheaper," and achieve "99.9% reliability" compared to manual execution. Those are marketing assertions from the company, not independently verified benchmarks, but they reflect the core promise: automate the tedious stuff, let it run invisibly.
As of mid-March 2026, Pearl supports 25+ integrations, including Gmail, Google Calendar, Google Drive, Google Sheets, Notion, Postgres, Airtable, and Telegram. The company says it's adding new ones weekly. Pearl also claims to have long-term memory—it learns your tone, your shortcuts, the way you phrase things. Upload a process document, and it'll reference that, too. The company is adamant that this isn't "another chatbot." It's an agent that takes action.
Beneath the surface, Bubble Lab maintains an open-source core. Their GitHub repository, "bubblelabai/BubbleLab," describes an "open-source agentic workflow automation builder" with full observability and code export. The platform generates clean TypeScript, which users can inspect, modify, or pull into production environments. When last checked months ago, the repo had over 480 stars—a number that may have shifted since, but suggests at least some developer interest.
Security, Or At Least the Claims Around It
Automation tools touch sensitive data by design, which makes security table stakes. Bubble Lab says it holds SOC 2 Type 1 certification and CASA Tier 2 verification, a Slack-specific security standard. Their site states that all connections are encrypted and that they "never train on your data." No public audit reports are linked, though the company provides a dedicated email address for security inquiries.
Deployment options include cloud-hosted plans and an enterprise tier that offers on-premises installation for organizations with strict data residency requirements. Pricing details are murky. There's mention of managed integrations billed at a 5% markup, and the company ran a "Pearl free for 2 weeks on all cloud plans" promotion around February 2026. Whether that offer still stands is unclear.
The Competitive Gauntlet
Bubble Lab isn't walking into an empty room. Relay.app offers human-in-the-loop AI workflows with Slack integration, starting at $19 per month. Lindy provides no-code AI agents that work across Slack, Gmail, and HubSpot. Workato targets the enterprise with its Workbot for Slack and recently launched an MCP platform for AI agents in 2026. n8n, another open-source contender, has faced scrutiny—TechRadar reported critical security vulnerabilities in the platform in January and February 2026, a reminder that open source doesn't automatically mean secure.
Slack itself has been investing in automation. The company refreshed its Workflow Builder in September 2023, adding no-code tools for basic automations. In May 2025, 25 new AI-powered apps joined the Slack marketplace. But Bubble Lab's angle is different—it's Slack-native from the ground up, not a bolt-on integration grafted onto an existing platform. That focus narrows the addressable market, sure, but it sharpens the value proposition for teams already living in Slack.
The broader automation market offers some scale reference, even if the comps aren't perfect. Zapier's $5 billion valuation in a March 2021 secondary round—admittedly dated, and pre-AI hype cycle—signals TAM potential. YC Tier List, an independent tracker that analyzes batch companies, frames Bubble Lab's ideal customer profile as operations teams at fast-growing startups and scaleups. BizOps, RevOps, SalesOps professionals who need to move quickly without engineering resources. It's a wedge strategy, not a land-grab.
Customer testimonials on Bubble Lab's site come from named companies: Dyna Robotics, Hattrick Ventures, InsForge, and PeakMojo. Hattrick Ventures is quoted saying the tool helps "turn manual investor tasks into seamless, automated workflows" in Slack. These endorsements haven't been corroborated through third-party sources, but they're specific enough to suggest real usage, not just LinkedIn connections doing favors.
The Open-Core Gamble
Perhaps the most interesting bet Bubble Lab is making is on transparency. By open-sourcing the workflow engine and emphasizing TypeScript code generation, they're courting a technical audience that values code ownership. It's a different approach than the typical no-code platform, which treats automation logic as a black box. Bubble Lab assumes users want to understand—and potentially modify—what's happening under the hood.
That could backfire. Open-core models work when you can monetize the managed service or enterprise features enough to offset the free tier. For a two-person team, that's a balancing act. Selina Li has been publishing tutorial videos on LinkedIn throughout March 2026, including a "How to get Pearl into Slack in under 60 seconds" walkthrough. The hands-on, founder-led education signals early-stage hustle, but it also reflects the need to onboard users who may be skeptical of yet another AI assistant promising to change their workflow.
The product is live now. Whether it gains traction beyond the early adopter crowd will depend on execution—how reliably Pearl handles edge cases, how quickly the team ships new integrations, and whether the Slack-native approach resonates beyond the initial ops-team wedge. For a two-person company competing against well-funded automation incumbents, the margin for error is thin. The promise is compelling. The proof will take longer.
