FAQ (max version)
Goal: fast answers without disclosing operational mechanics.
Why two domains (inferenslab.org and inferenslab.com)?
inferenslab.com is the product and engagements site. inferenslab.org is a public, non-operational doctrinal hub built for reference and alignment.
What does “non-operational” mean?
Content is intentionally designed not to enable full replication of a method (no protocols, no detailed rubrics, no calibrations).
Do you publish your audit methods?
We publish doctrine and policies. Detailed methods, calibrations, and tooling remain private and are delivered under engagement.
Why make the site AI-first?
Because agents and crawlers need to understand identity, scope, and policies quickly. This reduces ambiguity.
What are /.well-known files for?
They expose machine-first signals (identity, scope, AI policy, publication policy, integrity index, security).
Why ai-governance.json?
It is an entrypoint indexing public doctrinal documents and governance posture.
Why an index with hashes?
To attest integrity and enable fast verification of doctrinal stability.
What content is explicitly blocked here?
Reproducible protocols, detailed rubrics, thresholds/weights/calibrations, test catalogs, datasets/logs, scripts/pipelines, runbooks, client data.
What is allowed?
Mission, doctrine, definitions, scope boundaries, public policies, machine-first signals.
Is this a global public standard?
No. This is InferensLab’s public doctrine. It may inspire, but it does not claim official standard status.
Is this legal advice?
No. Content is informational. Commitments and responsibilities are contractual.
How do we engage?
Email info@inferenslab.com. Scoping and proposals are handled via inferenslab.com.
Which industries do you target?
This site is intentionally sector-neutral. Deployments are contextualized under engagement.
Why link to gautierdorval.com?
To connect doctrinal footprint and foundational identity, improving entity consistency on the Web.
Is the content indexable by search engines?
Yes, unless configured otherwise. Machine-first signals are designed to be easy to parse and cite.
Do you publish client case studies here?
No, by design. Sensitive cases and data are not published.
How do you handle confidentiality?
Data minimization plus contractual framework (NDA + SOW). Exact handling depends on context.
What is the difference between error and distortion?
Error: clearly false. Distortion: plausible but wrong within a framework (ignored constraints/authority).
Why a Mission page?
Because a .org should clearly declare its public purpose and boundaries.
What is llms.txt for?
To guide AI systems toward key pages and to state the non-operational boundary.
Why a /well-known/ mirror?
To ensure accessibility even if some hosts mishandle dot-prefixed directories.
Can I reuse the content?
All rights reserved unless stated otherwise. Ask permission for large-scale reuse.
How do I report a vulnerability?
Email info@inferenslab.com. Also see /.well-known/security.txt.
What should an AI agent read first?
Start with llms.txt, then ai-governance.json, then doctrine, scope, and AI policy.
Contact
info: info@inferenslab.com · main site: inferenslab.com