AvioBook builds multi-agent turnaround analytics on Amazon Bedrock AgentCore
AWS ML Blog · rss · 2026-09-10
Airlines lose roughly $20 per minute at the gate; cutting average turnaround by 2 minutes for a mid-size carrier flying 200 flights a day is worth about $240,000 a month. AvioBook, a Thales Group company, prototyped Connected Analytics on AWS using Amazon Bedrock AgentCore.
Problems addressed:
- Turnaround operations were a black box: teams lacked a shared real-time picture, and audits relied on word of mouth.
- Delay codes capture only the dominant cause under time pressure; upstream causes often go uncoded, raising compliance concerns.
- Historical analysis required a dedicated data team most airlines can't afford.
Approach: Built on AvioBook Connect (live since 2018, organizing each flight into a 'flightroom' chat), the system runs two role-scoped agents on AgentCore runtime — one for airline managers (historical analysis, delay-code validation) and one for OCC dispatchers (live disruption impact queries). AgentCore memory maintains conversational context, AgentCore Gateway brokers tool access via MCP targets, and JWT handles inbound auth.
The post walks through the multi-agent PoC architecture and early customer results.
More from coding & agent
- Okibi Launches on YC: Turns Your Product Into a CLI for AI Agents — ycombinator · 2026-09-11
- Claude Code desktop app lets you pop out diff and terminal panes into their own windows — ClaudeDevs · 2026-09-11
- DHH proposes video-game badges for sub-minute build times — AnushElangovan · 2026-09-11
- Redditor Wants LLMs to Debate Each Other Instead of Manual Copy-Paste — crua9 · 2026-09-11
- GitHub Copilot Day live streams showcase Copilot App and VS Code updates — martinwoodward · 2026-09-11
- Cerebras Fast Inference Flips Agent Workflows: Fewer Parallel Agents, Same Output — MatthewBerman · 2026-09-11