Method
Plan-and-execute agent
Plan first, then execute steps.
9 cases from 9 companies apply Plan-and-execute agent. 6 report results in production.
Typical tasks: Answer questions over data, Evaluate models, Explain a metric change, Measure impact without an A/B test and 5 more. Companies: DoorDash, Netflix, Databricks, Meta, Atlassian, Grab and 3 more. Often combined with: Text-to-SQL, LLM-as-a-judge, Semantic layer, Dense retrieval, Human in the loop and 19 more.
| Inside Vera, DoorDash’s Data Agent | food-delivery | 2026-09 | |
| A Human-Augmenting Agentic Workflow for Causal Inference | media-streaming | 2026-06 | |
| Anthropic + Adidas + Databricks: Unlocking 400 Hours of Productivity Weekly with Conversational AI | retail | 2026-06 | |
| How We Built Agent Mode for Genie | general | 2026-06 | |
A Corporate AI Agent at Nornickel![]() | ![]() | general | 2026-05 |
| Ranking Engineer Agent (REA): The Autonomous AI Agent Accelerating Meta’s Ads Ranking Innovation | adtech | 2026-03 | |
| How Rovo Deep Research works | saas-b2b | 2025-08 | |
| Streamlining RiskOps with the SOP agent framework | mobility | 2025-05 | |
| Practical Text-to-SQL for Data Analytics | social | 2024-12 |
