Pearl is Bubble Lab's Slack-native AI assistant for getting operational work done without leaving chat. Teams ask Pearl to execute workflows across Gmail, Jira, Notion, GitHub, and more than 25 connected tools, while developers can run the underlying Apache-2.0 workflow engine in their own infrastructure. It suits operations teams that live in Slack and developers who want observable, self-hostable automation.[1][2]
Quick verdict
Operations teams that live in Slack and developers who want self-hosted workflow infrastructure will get the most from Pearl. Pearl pairs a Slack-native AI operator with an Apache-2.0 TypeScript engine that traces every step and exports workflows anywhere. Teams that need generally available managed hosting today, or a turnkey no-code signup, should choose something else, since hosted Bubble Studio is in closed beta and self-hosting needs model keys and operator infrastructure. If your staff can operate infrastructure and hold model keys, choose Pearl for Slack-centered automation you control; if nobody will operate infrastructure, choose a managed-hosting automation platform instead.[1][2]
Key features
Slack-native AI operation Pearl works inside Slack, creating tickets, sending emails, querying databases, and running multi-step workflows from chat.[1][2]
Open-core Apache-2.0 engine The workflow execution runtime is Apache 2.0 licensed and production-ready on its own, while Pearl, the platform, and hosted infrastructure stay proprietary.[1][2][3]
Bubbles integration primitives Agents and integrations are built as ownable units the team can extend, share, or fork rather than opaque vendor nodes.[1][2]
Local Bubble Studio A visual flow builder starts locally in two commands and opens on localhost:3000 for building, editing, and running workflows.[1][2]
Step-level observability Execution tracing, logging, and introspection record what ran, which matters when autonomous work touches production systems.[1][2]
CLI project scaffolding The create-bubblelab-app command scaffolds a TypeScript project with templates such as basic and reddit-scraper.[1]
Exportable and embeddable workflows Flows export to run on the team's own backend or embed inside the team's own products.[1]
AI-native flow generation Natural language becomes working automation; generating flows with Pearl needs a provider key, with gemini-3.0-pro the documented default.[1][2]
Who will love it
Operations teams living in Slack Standup summaries, reports, lead enrichment, and ticket triage happen where the team already talks.[1][2]
Developers wanting a self-hosted typed engine TypeScript-native execution with tracing fits teams that distrust brittle JSON pipelines and opaque runtime loops.[1][2]
Teams outgrowing no-code prototypes The documented positioning targets groups whose n8n, Zapier, or Make setups stopped scaling on observability and control.[1]
Who should avoid it
Teams needing managed hosting now Hosted Bubble Studio is in closed beta behind an email request, so the supported path today is operating it yourself.[1][2]
Non-technical buyers expecting turnkey setup API keys, model providers, and infrastructure go through docs and configuration, not a guided signup.[1][2]
Solo users wanting one simple chatbot A person who only needs answers in chat will carry engine, studio, and integration machinery they never use.[1]
Alternatives and compatibility
Against n8n, choose the typed self-hostable engine Official docs name n8n as a prototype tool that falls short on scalability, observability, and developer control; choose Pearl when Slack-native operation plus a TypeScript engine you host is the requirement. This comparison follows the vendor's stated positioning and is editorial judgment, not vendor endorsement.[1][2]
Against self-hosted automation twins, weigh the Slack operator Where a team only needs engine-level automation without a chat operator, a twin open-core automator may carry less interface overhead; choose Pearl when the Slack operator is the point.[1][2]
Container teams get an extra incentive Teams already running container infrastructure can host the engine where their Docker and deployment practice already lives.[1][2]
Questions worth asking
What is the best Pearl alternative for non-technical users?
Is Pearl really no-code?
How much does Pearl cost per month?
Sources and editorial review
Source research. Reviewed on October 12, 2026. Written by Sanjib Sah. AI assisted; reviewed by a human editor.
We have not run or tested this product.
- Pearl discovery lead
official · Observed October 12, 2026
Discovery lead: Pearl as Slack AI assistant, bubblelab.ai, repo bubblelabai/bubblelab, Apache-2.0, stars/activity, 25+ integrations, YC backing.
- Bubble Lab official homepage
official · Observed October 12, 2026
Current homepage positions the Bubble Lab community/event intelligence layer; recorded as observed pivot, not identity conflict.
- bubblelabai/BubbleLab repository
official · Observed October 12, 2026
README: Pearl as Slack-native AI assistant/operator interface; open-core engine scope; stars/forks; managed platform vs self-hosted engine; local/CLI setup; provider-key requirement; proprietary exclusions.
- Bubble Lab documentation intro
official · Observed October 12, 2026
Apache 2.0 release; explicit n8n/Zapier/Make comparison; TypeScript-native, tracing, Bubbles ecosystem, self-hostable runtime, AI-native; hosted studio closed beta.
- BubbleLab LICENSE (Apache-2.0)
official · Observed October 12, 2026
Full Apache License Version 2.0 text in LICENSE.txt.
Compared with these products
No explicit proprietary baseline is recorded.