Skip to content
0727
InsightsAcademyCasesEventsSign in
中文/EN
Cases/Meta
← Back to cases
← Previous15 / 127Next →

Meta

Expert-knowledge agent and verified correction loop for a specific compliance domain
First-party enterprise source Internet / enterprise compliance Internal enterprise AI deployment ⚙ Technical reference Deep case | Key delivery chain is substantially documented Evidence level A Internal system for a specific domain / continuing iteration
A Evidence level
An Organizational Second Brain: Building an AI That Learns From Experts
Publisher: Engineering at Meta · Company engineering blog · Direct case-level source
Claim origin: Company engineering-team disclosure · Independent verification: No · Accessed: 2026-09-23
Evidence level measures whether a source can be located and reviewed; it does not mean vendor-reported claims were independently audited.
Deep case | Key delivery chain is substantially documented 12 / 12
Business context2 / 2
Transformation workflow2 / 2
Technical workflow2 / 2
Human roles & governance2 / 2
Measured outcomes2 / 2
Source traceability2 / 2
All six dimensions meet the current completeness threshold.
Meta separates specialist knowledge files from analytical recipes and converts expert corrections into minimal, reviewable changes tested by blind replay and regression suites before a human-approved merge.
Business problem

In a specific compliance domain, experts must synthesize organizational positions, historical decisions, and external material. Important reasoning lives in scattered documents or in people's heads. Basic RAG still requires the model to infer rules from raw chunks on each run, making assessments slow, inconsistent, and difficult to improve safely after a human correction.

Solution

Meta distilled more than 200 files into positions, vocabulary, routing rules, and gateways, using YAML dependencies to make the knowledge graph traceable. Frequently used dense knowledge lives in a structured wiki, while sparse references remain searchable through RAG. Composable recipes specify analytical steps separately from domain facts. Expert feedback goes through root-cause diagnosis, minimal file edits, blind replay, regression tests, and human-reviewed pull requests.

Technical architecture & production workflow
Step 01
Distill internal and external material into 200+ position, terminology, routing, and gateway files
→
Step 02
Use YAML depends_on and referenced_by metadata to track file applicability and dependencies
→
Step 03
Keep dense frequent knowledge in a structured wiki and retrieve sparse references with semantic or lexical RAG
→
Step 04
Route analyses through top-level and downstream recipes, loading only relevant knowledge by stage
→
Step 05
Expose intermediate analyses to experts at checkpoints and escalate genuine ambiguity
→
Step 06
Classify expert corrections as knowledge gaps, recipe defects, or ambiguity before proposing minimal edits
→
Step 07
Require independent review, deterministic lint, blind targeted replay, regression tests, and expert-reviewed PRs
Key technology & infrastructure components
Structured knowledge wikiBidirectional YAML dependency graphRAG retrievalComposable recipe workflowsExpert checkpoints and escalationCorrection-compilation pipelineDeterministic lintBlind replay and regression tests
Human roles & accountability

Domain experts review intermediate analysis at checkpoints, redirect it, and resolve genuine ambiguity. Automated updates produce proposed diffs with an audit trail, but expert approval is required before changes land. Independent reviewing agents and deterministic structural checks help catch contradictions without replacing expert accountability.

FDE delivery actions
  • Select a compliance domain with costly, repeated expert judgments
  • Distill reviewable organizational positions rather than indiscriminately searching every document
  • Separate domain facts from analytical procedures, with explicit routing and dependencies
  • Design expert checkpoints and ambiguity escalation into the workflow
  • Diagnose each correction before applying a minimal knowledge or procedure fix
  • Gate changes through independent checks, replay, regression tests, and expert approval
Reusable delivery patterns
  • Encode expert knowledge as versioned, auditable facts and procedures, not only vectors
  • For high-stakes judgment, engineering the correction loop may be more valuable than generating more answers
  • Turn each expert correction into a versioned change and a regression example
  • Internal efficiency claims should be read alongside missing sample and baseline details
Business outcomes & delivery results

Meta Engineering reports that after 3 sprints over about 6 weeks, individual assessments moved from days to minutes. Recipe-driven stages cut tokens per turn by about 80%; the team reports zero regressions across improvement cycles. Sample size, detailed timing baselines, test counts, and independent audit are not published.

Evidence boundaries & verification notes
  • This is a specific compliance-domain system, not the separate Meta Second Brain project for a broad knowledge-worker population.
  • The days-to-minutes claim has no public sample size or statistical baseline; zero regressions has no published test count.
  • The article does not establish company-wide rollout, so the case is marked as a domain-specific internal system.
Primary source: An Organizational Second Brain: Building an AI That Learns From Experts ↗
Traceable does not mean independently audited
Open primary public source
0727.ai · Trusted agents, built together.Case research: FDE-case-library ↗ · MIT