eltPulse

Pulse AI

Pulse AI is eltPulse's Claude-powered assistant for pipeline design. It lives in the pipeline builder create flow and as a floating widget (plus an inline bar on the visual canvas). Describe what you want to load — source, destination, optional dbt transform, run slices — and Pulse AI generates a save-ready definition with inline credential fields where needed.

Visual pipeline design with AI-assisted editing — on any warehouse, tied to real dlt, Sling, and dbt codegen you can export to Git.

Who can use it

Workspace roles with pipeline write (owners, solo users, and member invites) can save AI-generated pipelines. Viewers and catalog-only roles get a read-only message.

What Pulse AI generates

  • Core pipeline — name, source type, destination type, connector-specific sourceConfiguration, and Python/YAML preview.
  • Canvas patches — when editing on the canvas, Pulse AI can add transform nodes, wire edges, and update inspector fields — not just greenfield scaffolding.
  • Post-load dbt (EL+T) — sets sourceConfiguration.dbt with package path, Git URL, schema, branch, and model selector.
  • Linked dbt project — calls list_dbt_projects and sets dbtProjectId to link a workspace project from /catalog/dbt.
  • Python / SQL post-transform — optional post_transform block for scripts after load.
  • Workspace context — list pipelines, connections, and catalog assets to answer setup questions.

Human-in-the-loop save

  1. Review the generated code preview in the chat.
  2. Click Looks good to reveal save options.
  3. Fill inline fields (e.g. GitHub org/repo) or skip and edit in the form builder.
  4. After save, open the pipeline in the form builder or visual canvas.

Example prompts

  • Load GitHub issues and PRs into Snowflake
  • Sync HubSpot contacts to BigQuery with a dbt staging model after load
  • Add a dbt transform node between Stripe and Snowflake on the canvas
  • Backfill HubSpot with run slices for last 30 days
  • What workspace dbt projects do I have?

MCP servers

Register Model Context Protocol servers at MCP servers to extend Pulse AI with custom tools (HTTP or stdio transports).

API

POST /api/elt/ai-assistant accepts { messages: [{ role, content }] } and returns message, optional savePayload, requiredFields, and codePreview. Requires session auth (or workspace API key where enabled) and write permission.

Pipelines & canvas · dbt transforms · Catalog