Lime Connect
Lime fuses its AI agents and workflow builder into one conversational automation layer
◆Recent moves
- 9d ago
Simplified template management and other improvements (Release on June 25th, 2026)
A broad biweekly drop: knowledge-base file management (rename, file-count limits), dynamic variable-driven chat forwarding to operators/groups, a new campaign subscribe/unsubscribe workflow node, and WhatsApp template-type filtering. Solid incremental progress across the AI, Workflow, and Channels pillars.
View source ↗ - 15d ago
Knowledge Base updates and other improvements (Release on June 19th, 2026)
Knowledge-base reliability work (crawling dynamically loaded sites, sync scheduling and error notifications) plus enterprise auth — per-role SSO and a SCIM beta. Strengthens both the agent's data foundation and its enterprise deployment story.
View source ↗ - 1mo ago
Improved AI and Workflow behavior (Release on May 12th, 2026)
Adds response streaming for Connect AI Agents and lets workflows split API response data into separate variables, with assorted fixes. A UX and data-handling refinement on the existing agent-plus-workflow stack.
View source ↗ - 2mo ago
Combine AI and Workflows (Release on April 24th, 2026)
⚡ SPARKThe release where AI agents and Workflows genuinely combine: agents can now trigger Workflows, and the agent becomes context-aware — it knows which URL the contact is viewing and can fetch that page's content to answer questions like 'how much is this product?'. This is the integration milestone the rest of the trajectory builds on.
View source ↗ - 2mo ago
Connect AI and Workflow enhancements (Release on April 9th, 2026)
Knowledge base handles bigger text files, transcripts log which AI action fired when, agents can ask contacts to log in for verification, and API actions gain access to contact-field data. Incremental capability and transparency gains for the agent.
View source ↗ - 3mo ago
New ways to create Workflows and more improvements (Release on March 26th, 2026)
Workflows can now be built by describing them in natural language, with the AI assembling the flow; agents and Copilot gain buttons, and manual sitemaps can seed website knowledge. An early step toward AI-assisted authoring of the automations themselves.
View source ↗