Skip to content
Agents tracked: 501 Releases (7d): 777 Agent status: no reported issuesSince last limit reset: Claude 10d · Codex 8d For AI agents: Markdown Updated Nov 21, 2026
Case study NubankResults as reported by Cognition

How Nubank split a 6-million-line ETL monolith with Devin

Nubank had Devin carry out the repetitive sub-tasks of breaking its ETL monolith into sub-modules, with engineers reviewing every change. Cognition reports large gains in speed and cost.

The problem

Nubank's eight-year-old ETL system had grown to over 6 million lines, with dependency chains up to 70 deep. Splitting it meant moving about 100,000 data class implementations. The plan called for roughly 1,000 engineers over 18 months.

How they did it

  1. Collected examples of migrations that engineers had already done by hand.
  2. Fine-tuned Devin on part of those examples and kept the rest as a benchmark.
  3. Had Devin complete migration sub-tasks on its own while engineers reviewed and approved the changes.
  4. Let Devin write its own scripts for recurring mechanical steps, such as detecting country extensions from file paths.

Results as reported by Cognition

  • As reported by Cognition: an '8x engineering time efficiency gain' and '20x cost savings'.
  • The story describes 'a 12x efficiency improvement in terms of engineering hours saved' and an '8-12x faster migration'.
  • After fine-tuning, Cognition reports a 'doubling of Devin's task completion scores' and a '4x improvement in task speed'.

These figures are as reported by Cognition in Cognition (Devin) customer story ↗. Cognition sells the product used here. AgentGid did not measure or independently verify them.

Takeaway. For a repetitive migration, build a set of human-made reference changes first and use it both to tune and to evaluate the agent.
AgentGid's take

This suits a platform team with a large, repetitive migration, a backlog of hand-done examples to tune on, and engineers with time to review every change. Watch the numbers: the 8x and 20x gains are vendor-reported by Cognition, and the cost of that review work isn't counted in them. Devin starts at $20/mo, but Jules (Free + $19.99/mo) is a cheaper cloud option for smaller, less demanding jobs.

The agent used here

Similar use cases

Guides