Agentic Workflows. The new way of analytics.

Configure your data sources, add AI agents in a topology, and set a goal. Agentic Workflows run on schedule and deliver analyses to your app and inbox when they're ready. So you stop chasing dashboards and get insight when it matters.

API
group
4o-mini 11 tools
4o-mini 4 tools
4o 11 tools

What is an Agentic Workflow?

An Agentic Workflow is an analytics pipeline you configure once: you choose where your data lives, which AI agents run in which order, and what "good" or "bad" looks like. After that it runs on your schedule and delivers results to the app and your inbox.

  • Connect your data once: files, S3, APIs, or databases.
  • AI steps that do the work: cleaning, profiling, analysis, visualization. Configure the ones that you need.
  • Set guardrails to stay in control. Max steps and timeouts keep results predictable; you'll always see what each agent did.
  • Data flows in a clear order through your agents.
  • Run on a schedule, or on demand.
  • Every run is saved as an analysis in the app. Add optional email and insights land in your inbox when they're ready.

Full Transparency

See exactly what happened

Every Agentic Workflow run produces a detailed execution trace. See each agent's decisions, routing logic, and outputs.

Trace for Agentic Workflow: Analyse Pipeline

View execution logs and debugging information for this analysis

Filter 142 All origins
executor

Branch 1c4e57c6-76f2-45d1-8331-39edd8c31537 step 1: agent=EDA success=True response_len=258 preview={"action": "done", "action_input": {"summary": "The comprehensive data profile has been successfully created and documented. The next analysis agent c…

18:49:55,789
executor

Branch 1c4e57c6-…39edd8c31537 validation: is_valid=True action_type=final_response error=None

18:49:55,789
⚘ routingrouter

Goal-based routing evaluation response: {"next_agent": "Analysis", "reason": "The EDA has been completed successfully, and the next step is to perform the analysis based on the insights gathered."}

18:49:56,557
⚘ routingrouter

Goal-based routing: EDA -> Analysis

18:49:56,557
executor

Branch 1c4e57c6-…39edd8c31537 routing: should_continue=True next_steps=['Analysis'] final_response=None

18:49:56,557
executor

Branch 1c4e57c6-…39edd8c31537: Transitioning from EDA to Analysis

18:49:56,557
agent_builder

Analysis Agent iteration 1

18:49:56,559
agent_builder

Analysis API response: finish_reason=tool_calls, has_content=False, tool_calls=1

18:49:57,490
agent_builder

Analysis executing tool: read_blocks args={analysis_id='76f7bd2d-0672-…'}

18:49:57,490

Full visibility

Every agent step, routing decision, and data handoff is logged and visible in the trace.

Real-time streaming

Watch the trace populate live as agents execute, or review completed runs at any time.

Debug with confidence

Pinpoint exactly where and why a result was produced. No guessing, no black boxes.

Why datalyft

What makes Agentic Workflows unique

Agentic analytics you build once. Insights you get when it matters.

Agentic, not just automated
Multiple specialized agents work in a defined topology. You get structured Agentic Workflows, not a single black box.
Configure once, run on schedule
Set daily, weekly, monthly, or custom cron; timezone-aware. No need to remember to run the report; insights are generated and delivered automatically.
Insights delivered to you
Analysis is always stored; optional email on completion with summary and preview. So "insights come to you" is literally true: inbox and app.
Transparency and control
Execution trace, step-by-step view, saved analyses. You can verify how a result was produced. Real data sources: File Upload, S3, API, Database. One place to configure Agentic Workflows, set schedules, and see run history.

In 4 steps to actionable insights

  1. 1

    Connect your data

    Connect the sources you already use: files, APIs, or databases.

  2. 2

    Configure your Agentic Workflow

    Add agents from templates and connect them in a topology. Set your data sources and goal. No code required.

  3. 3

    Set when it runs

    Run now or on a schedule: daily, weekly, monthly, or your own rhythm.

  4. 4

    Get results in the app and your inbox

    When the run finishes, the analysis is saved in the app. A summary lands in your inbox when relevant insights are found.

Get access

Join teams who get insights when it matters.

  • Always-on analytics that works for you.
  • Configured once. Continuous Insights.

Early access

Be among the first to build Agentic Workflows.

Founding Users Benefit

Get access to the private beta, direct input into our roadmap and custom integrations.