Suraj Gupta

Suraj Gupta’s 2026 AI Engineer World’s Fair biography described him as a software engineer at Warp working on agentic developer experience and AI terminal workflows, including Warp’s agentic development environment. In his presentation, “How Software Factories Improve Themselves,” he identified his role at the time as leading harness development at Warp and demonstrated three approaches to improving automated software workflows: evolving agent skills, persistent memory, and model routing.

Using examples from Warp’s software factory, Gupta showed how an outer-loop agent examines triage decisions and human feedback to propose skill updates through Git-tracked, human-reviewed pull requests. He also presented persistent memory in Warp’s Oz cloud agent platform, where agents retain facts and outcomes from earlier investigations, helping a Sentry agent reuse prior root-cause findings. These memories work across harnesses and can be versioned, edited, deleted, and traced to source runs. His model-routing examples covered automatic selection and custom rules that assign task classes to suitable models to manage cost. He described Warp’s internal evaluations comparing models on tasks, reporting efficient GLM performance on UI work, and outlined planned customer-facing evaluations tailored to individual workflows.