B2B questions and answers (draft)
Status: DRAFT — pending counsel review (Linear AIP-9).
Not for publication or customer use. Dev-only artifact for counsel and sales enablement (Linear AIP-122; owner answer 7, 3 Nov 2026).
Companion drafts: B2B Privacy Policy (draft) (AIP-121); B2B Terms of Service (draft) (AIP-7).
These answers describe the intended self-serve B2B API at launch. Switches that expose signup or scans to production customers stay off until counsel and product gates clear (AIP-9).
Who is this for?
AI Powered Social Scan sells a business-to-business (B2B) platform. Customers are companies and other organizations (not consumers buying a personal app) that want to review public social-media activity for lawful business purposes — for example hiring, vendor screening, marketplace trust and safety, or similar compliance workflows.
At launch we accept self-serve signup only for companies registered in the United States or Israel. Representatives of those companies create an account, complete a short company questionnaire, and (when approved) use our API or developer portal to run scans.
We do not offer the B2B product to individuals acting for personal curiosity, covert employee monitoring, or undisclosed law-enforcement use. Our access rules may block or require manual review for high-risk declared use cases (see company check).
What do you scan?
When a Customer submits a scan, we collect and analyze content from publicly available social profiles and posts the Customer points us to — typically profile URLs for major networks supported by the platform (including X, Facebook, Instagram, TikTok, and LinkedIn, subject to what is publicly visible on each network).
Our pipeline generally:
- Retrieves public profile metadata and public posts (and related public media where applicable).
- Transcribes or processes public audio/video needed for review.
- Ranks and scores posts for relevance using automated models.
- Produces a report (JSON and/or interactive HTML/PDF-style outputs) with risk flags and supporting evidence drawn from public content.
What we do not do by design:
- We do not break into private accounts, bypass login walls, or collect from non-public sources outside the Customer-directed job.
- B2B reports exclude certain consumer-only report elements (for example Snitch payloads and sentiment/topic color overlays on word portraits) — see product spec AIP-276.
Customers choose who to scan and why; they remain responsible for having a lawful purpose and appropriate notices to the people they scan (see consent and notice).
What about consent and notice to the person being scanned?
Short answer: The Customer (your organization) is responsible for telling scan subjects what you are doing and for having a valid legal basis under your jurisdiction. We provide the screening tool; you make the decision to run a scan.
At signup we ask representatives to answer honestly about:
- Declared use cases (for example hiring vs marketplace safety).
- Countries where scan subjects are located.
- Whether your organization already has subject consent and whether subjects are informed about the screening — you answer
yes,no, orunsurewhere applicable.
Those answers feed our company check (rules engine). They are not a substitute for your own privacy notice, consent forms, or adverse-action process where the law requires them (for example FCRA-style hiring workflows in the United States).
If someone contacts us about a scan your organization ran, our Privacy Policy draft describes the intended path: we generally ask them to contact you first, and we help route or facilitate requests when we act as a service provider processing on your instructions.
Counsel must finalize customer-facing consent and notice language in your Terms, Privacy Policy, and Data Processing Agreement (AIP-9).
What is the company check, and when can we scan?
Self-serve signup is open, but scanning stays locked until your organization passes an automatic company check (also called the access questionnaire and traffic-light model — AIP-91).
How it works:
- A representative submits the questionnaire (
POST /v1/access/applications) with company legal name, website, country of registration, contact details, use cases, expected scan targets, subject countries, and consent/informed flags. - A counsel-configured rules engine evaluates the answers and assigns a door:
- Green: access is granted automatically, subject to verification steps (bot/abuse checks and company validation) — you receive a tenant account, a one-time API key, and starter credits; scans unlock subject to billing and technical gates.
- Yellow: your application is held for operator review before scans unlock.
- Red: the declared use is prohibited or incompatible with the Service in v1 (for example personal curiosity, covert employee monitoring, or law-enforcement use). Self-serve access is declined; an operator may review the application, and you may receive an explanation or next steps at counsel’s discretion (Terms draft §4.1, §6).
- You can poll application status until the door is resolved.
This check is about eligibility and abuse prevention, not about approving each individual scan. After you are green (or approved from yellow), you choose profile URLs to scan via the API, subject to your balance, price cap, webhooks, and rate limits.
Nothing in this draft promises a specific turnaround time for yellow-door review; operations SLAs are a counsel/ops decision.
How does pricing work?
Pricing is usage-based in USD credits, not a flat unlimited subscription at launch.
| Topic | Launch intent (see AIP-33 / AIP-34) |
|---|---|
| Credit unit | 1 credit = $1.00 USD (ledger stores cents). |
| Signup credit | A free signup credit once your organization passes the company check (green or approved path). The amount is configured in Mission Control, not fixed in this draft. |
| Paid top-ups | Credit packs via Stripe Checkout when billing is enabled (AIP-36). |
| Per-scan meter | On successful scan completion: $1.00 + $0.008 × (number of posts analyzed), charged in cents, never more than $15.00 per scan. |
| Price cap (per scan) | $15.00 maximum per scan; the cap is reached at 1,750 posts analyzed ($1.00 + 1,750×$0.008). A single scan covers at most 3,500 posts. |
| Hold at create | When you start a scan, we hold your full active cap until the job finishes; unused hold is released; on success we debit the metered amount. |
| Minimum to start | Available balance must be at least your cap; otherwise the API returns a payment-required response. |
| Failed scans | Hold is released; no metered debit on failure/cancel paths per ledger design. |
Example: a scan that analyzes 100 posts meters $1.00 + 100×$0.008 = $1.80 if it succeeds (still well below the $15.00 per-scan maximum).
We may block scan creation if projected pipeline cost would exceed your cap or platform safety limits — see API documentation. Tax, invoices, and refund policy belong in Terms and billing disclosures (counsel + AIP-36).
What quality assurance do you perform on reports?
Before customer-facing launch, the owner’s default QA policy (dev configuration) is:
- Manual QA on the first three scans for each new Customer organization.
- After that, 10% spot checks on scans (sampling — not every flag on every report).
QA reviewers are platform operators, not your end customer. QA may review flagged content, transcripts, and report assembly for accuracy and policy compliance. Customers should still apply their own human review before making decisions about individuals.
Counsel should confirm whether this sampling requires extra notice in your DPA or privacy materials (Privacy Policy draft §4).
Which regions do you support?
Customers (who may sign up self-serve): companies registered in the United States or Israel only at launch. The questionnaire captures country_of_registration; other countries are out of scope for self-serve until we announce expansion.
Scan subjects: Customers declare where subjects are located; some use-case and jurisdiction combinations trigger yellow-door review or duties (for example hiring in the US or EU/UK) in the rules matrix — counsel maintains the matrix.
Where data is processed: production infrastructure is built around AWS Europe (Frankfurt, eu-central-1) as the primary region, with backup and disaster-recovery patterns documented internally (AIP-124 / AIP-127). This is not the same as “data never leaves the US” — counsel must align marketing language, DPAs, and transfer tools with reality.
Region setting vs. hosting location: the “region” setting (dev only for now) is the Customer’s company country (US or Israel). It decides which launch rules, legal terms, and review duties apply to that Customer. It does not move where data is hosted: hosting stays in the primary AWS region above. Counsel validates the cross-border transfer story.
How do I get started (high level)?
- Submit the access application (company questionnaire).
- Pass the company check (green, or yellow then approval).
- Configure webhooks and store your API key securely (shown once).
- Fund credits (signup grant plus optional packs).
- Call
POST /v1/scanswith public profile URLs and a purpose attestation.
Detailed API steps live in the developer portal and OpenAPI spec — not in this Q&A.
Where this Q&A fits with Terms and Privacy
| Document | Role |
|---|---|
| Terms of Service (draft) | Binding contract: eligibility, permitted uses (rules matrix), acceptable use, warranties, liability, billing — counsel to publish. |
| Privacy Policy (draft) | How we handle personal data (account vs scan subjects), retention, rights, subprocessors. |
| This Q&A | Plain-language sales and onboarding FAQ; if anything here conflicts with Terms or Privacy after counsel review, Terms and Privacy win. |
Appendix A — AIP-122 / owner answer 7 coverage map
| Task requirement | Q&A section |
|---|---|
| Who it is for | Who is this for? |
| What we scan | What do you scan? |
| Consent and notice | What about consent and notice… |
| Company check | What is the company check… |
| Prices | How does pricing work? |
| QA | What quality assurance… |
| Regions | Which regions do you support? |
| DRAFT + counsel gate | Header banner |
| Alignment with Terms and Privacy drafts | Where this Q&A fits… + Appendix B |
Appendix B — Consistency checkpoints (Terms / Privacy drafts / product specs)
These statements are intentionally aligned with docs/legal/drafts/b2b-terms-of-service.md, docs/legal/drafts/b2b-privacy-policy.md, and approved P4 specs. Counsel should reconcile wording across all customer-facing docs.
| Topic | Q&A statement | Terms / Privacy / spec anchor |
|---|---|---|
| Launch countries | US + Israel self-serve signup | Terms §3.2; Privacy §5; questionnaire |
| Company check | Green / yellow / red; scans locked until resolved | Terms §4.1, §6; Privacy §3.1; AIP-91 |
| QA sampling | First 3 scans per new customer; then 10% spot checks | Terms §4.4; Privacy §4 |
| Metering | $1.00 + $0.008×post_count, max $15.00 per scan (cap at 1,750 posts); scan size up to 3,500 posts | AIP-33; Terms §8 points to public pricing docs |
| Signup credits | Free signup credit on company-check approval (amount via Mission Control) | Terms §8 (welcome grant on green-door approval); AIP-33 / AIP-41 |
| Controller vs processor | Customer controls scan subjects; we process on their instructions | Terms §9; Privacy §2 |
| Public content only | Public profiles/posts | Terms §5; Privacy §3.2 |
Document version: draft 2026-11-03 — repo path docs/legal/drafts/b2b-q-and-a.md.