Pipeline teams

Research assistants for literature review, experiment execution and report preparation.

A research pipeline evaluated on scientific implementation and manuscript tasks.

Planner and coding agents built into Atlassian's internal Jira that take a work item to a pull request, with an engineer approving each step.

An Italian online daily that is researched, written, verified and edited by five AI agents and goes live without prior human review.

Spotify's internal agent setup for code changes: an interactive agent scopes the task, a background coding agent opens the pull request, and an LLM judge checks the diff.

Non-agentic 3-stage pipeline — 40.7% Lite / 50.8% Verified with Claude 3.5 Sonnet.

Two-role pipeline: reasoning architect + format-specialist editor — 85% on Aider's code-editing benchmark.

Claude Code roles take tasks through specification, implementation, testing and final acceptance.

Internal fleet-scale Java 8/11→17 upgrades — >4,500 dev-years saved, $260M.

5-role sequential pipeline — 22,949 tokens, 148s per software task.

Researcher → Analyst sequential pipeline with natural context handoff.

An internal on-call copilot answering questions from Uber documentation.

Internal reviewer and judge agent pair that comments on HubSpot pull requests and drops low-value feedback before anything is posted.

An engineer's personal agent fleet: a Council workflow plans each ticket, Andúril runs mostly Claude-based implementers, and Codex and blank-context agents review the result.

An early personal-blog setup for Japanese writing, English translation and editorial support.

Planner + executor pool + per-step verification — vendor-claimed 77.6% on GAIA.

Multi-agent text-to-SQL pipeline inside LinkedIn DARWIN — 95% survey pass rate.

39.33% on SWE-bench Lite (lab-page claim).

5-agent SOP-encoded pipeline — 124.3 tokens/LoC, executability 3.75/4.

Sam, Barry, Karen and a CRM agent answer leads, check calls for compliance, route complaints and keep CRM records current for Send Payments.

Specialist analysis agents and domain judges that propose and vet changes to Shopify's product taxonomy before the taxonomy team reviews them.

Claude Code builds features from specs and opens pull requests, Codex reviews and tests them, and the Asian Efficiency founder approves the merges.

A one-person Tucson local news site where AI models write the morning brief, preview council agendas, draft meeting reports and voice a podcast and a Spanish newscast.

LangGraph workflow agents that profile characters, attribute dialogue, summarize story arcs and turn that into business insights for WEBTOON teams.

Bilingual support pod: frontline resolution plus localization.

Plan → Write → Edit content pipeline with brand-voice enforcement.