{"type":"post","id":"how-shopify-turns-open-models-into-task-experts-edc9e2","slug":"how-shopify-turns-open-models-into-task-experts-edc9e2","title":"How Shopify turns open models into task experts","description":"Shopify’s task-expert loop makes a narrow agent workload—not a general model leaderboard—the unit of optimization.","retrieval_nugget":"Shopify’s task-expert loop makes a narrow agent workload—not a general model leaderboard—the unit of optimization.","published_at":"2026-10-21","updated_at":"2026-10-21","record_date":"2026-10-21","date_kind":"published_at","topics":["open-models"],"source_urls":["https://shopify.engineering/open-model-task-experts"],"schema_version":"newruntime-agent-readable-v0.2","stable_id":"post:how-shopify-turns-open-models-into-task-experts-edc9e2","status":"published","visuals":[],"editorial_provenance":{"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:2303abe9f87f12450824a7fc09e5d01f5eaab3975d0c55ec9c64c4c43082560e","reviewed_at":"2026-10-21T12:05:20.638Z","source_evidence_count":1,"verified_claim_count":3,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":3,"implication_count":1,"watch_condition_count":1,"related_record_count":0},"analysis":{"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Shopify’s task-expert loop makes a narrow agent workload—not a general model leaderboard—the unit of optimization.","observed_facts":[{"text":"Shopify reports Nemotron 3.5 Lightning scored 74.5 on its GraphQL judge after six loops, compared with 73 for the frontier-model baseline.","source_urls":["https://shopify.engineering/open-model-task-experts"]},{"text":"Shopify reports self-healing repaired 59,535 GraphQL conversations over 86 nightly runs, or 69% of attempted failures.","source_urls":["https://shopify.engineering/open-model-task-experts"]},{"text":"Shopify reports a 0.87-second median response in replay testing versus 6.50 seconds for the frontier API, with about 90% lower serving cost.","source_urls":["https://shopify.engineering/open-model-task-experts"]}],"mechanism":"A calibrated task judge scores candidate changes and checkpoints, while a self-healing process converts some failed conversations into training examples.","why_now":"Shopify is describing a deployed workflow in which open-model post-training, evaluation, and self-hosting are linked to its commerce assistant rather than presented as isolated experiments.","implications":["For bounded, high-volume subagent jobs, a continually evaluated open model can be a cheaper and faster production component than routing every step to a frontier API."],"evidence_boundary":"These are company-reported results on Shopify workloads; they do not establish independent replication, generalize to other tasks, or show the full engineering and GPU costs of operating the loop.","watch_conditions":["Whether Shopify reports sustained quality and cost across additional task experts and workloads."],"related_records":[],"new_branch_reason":"No prior New Runtime record was supplied; this creates a distinct analytical record on Shopify’s production task-expert training and evaluation loop."},"editor":{"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]},"routes":{"html":"https://newruntime.com/posts/how-shopify-turns-open-models-into-task-experts-edc9e2/","markdown":"https://newruntime.com/posts/how-shopify-turns-open-models-into-task-experts-edc9e2.md","json":"https://newruntime.com/posts/how-shopify-turns-open-models-into-task-experts-edc9e2.json"},"related_patterns":["ai-engineering-evaluations-are-splitting-into-task-specific"],"next_reads":[{"type":"topic","path":"/topics/open-models/","url":"https://newruntime.com/topics/open-models/","title":"Open Models","reason":"Explore the open-models topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/open-models.md","json_url":"https://newruntime.com/topics/open-models.json"},{"type":"collection","path":"/posts/","url":"https://newruntime.com/posts/","title":"Field Notes","reason":"Open the latest editorial analysis.","media_type":"text/html","json_url":"https://newruntime.com/posts.json"},{"type":"collection","path":"/source-ledger.json","url":"https://newruntime.com/source-ledger.json","title":"Source Ledger","reason":"Inspect the public source evidence graph.","media_type":"application/json"}]}