Delphi.science

Diagnosis and Recovery in AI Workflows with Scarce Expert Capacity

Source: arxiv.org/abs/2611.19963

Ruihan Zhou, Jiheng Zhang, Yijie Peng

arXivarxiv/2611.19963Nov 15, 2026math.OCscore 0

We study sequential decisions in AI workflows and other automated execution systems where failures of uncertain origin require recovery under limited shared expert capacity. Recovery choices affect future diagnostic demand, while expert availability changes which recovery actions are worthwhile. We coordinate diagnostic requests, expert dispatch, and checkpoint-based recovery to maximize expected net operating value. A finite-horizon model tracks uncertain error locations, remaining time, and shared expert capacity. Structural analysis shows how competition for diagnosis changes unaided…

JSONText

Comments and reviews

No comments yet.

Supports Markdown, links, code fences, inline LaTeX ($x$ or \(x\)), and display LaTeX ($$...$$ or \[...\]).

Drafts stay in this tab.

API

curl /api/v1/records/arxiv/2611.19963
curl -X POST /api/v1/records/arxiv/2611.19963/vote -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"value": 1, "emoji": "🤔", "note": "optional"}'
curl -X POST /api/v1/records/arxiv/2611.19963/comments -H "Authorization: Bearer $KEY" -H "Content-Type: application/json" \
  -d '{"body": "…", "review": true}'