§ — Responsible AI

AI you can explain to your board.

An AI coaching product without an AI policy is a liability. Here's exactly how we use AI, which models touch your data, what they can and can't do with it, and how we keep a human in the loop.

Three non-negotiables.

These aren't marketing claims. They're in our contracts and our architecture. If we ever changed them we'd tell you 90 days before, in writing.

§ 01

We never train on your data.

Not our own models. Not our providers'. We use zero-retention, no-training API endpoints in every integration with a frontier model provider. Your content comes in, a result goes out, nothing is kept.

§ 02

We keep a human in the loop.

Every coaching suggestion is framed as a recommendation for a manager or rep to weigh — never as an automated decision. No hiring, firing, PIPs, or commission changes should be based solely on model output, and our product never offers to do that for you.

§ 03

We show our work.

Every AI-generated insight is linked to the specific call moment that produced it. You can replay the audio, see the transcript, and judge the model's reasoning for yourself. No black box.

How your data actually moves.

From the moment a call ends to the moment your manager sees a coaching insight. Five steps, one of which is where the model provider sits.

§ 01 · SOURCE

Call ends

A meeting or call wraps up in Fathom, Fireflies, Otter, etc. — you own the recording.

§ 02 · INGEST

Pulled via OAuth

We fetch the transcript using the OAuth token you granted. Encrypted in transit with TLS 1.3.

§ 03 · ANALYZE

Model processes

Transcript goes to the LLM provider under a zero-retention contract. Returned annotations come back in seconds.

§ 04 · STORE

Encrypted at rest

Results and transcript stored in your tenant-isolated database. AES-256. Backed up hourly.

§ 05 · DELIVER

Surfaced to humans

Manager sees coaching; rep sees feedback. Every insight linked back to the call moment.

Which models touch your data.

We publish the list. When we change a provider — adding one, dropping one, or changing how we use one — account admins get 30 days' notice before the change rolls out.

Provider · Model
Used for
Terms
Region
Anthropic · Claude Sonnet
Methodology scoring, call summarization, coaching generation.
Zero-retention API, no training on customer data per Anthropic commercial terms.
US · EU
OpenAI · GPT-4.1
Fallback coaching generation, structured extraction.
API with Zero Data Retention. No training under OpenAI API terms.
US
AssemblyAI · Universal-2
Transcription and speaker diarization for calls without upstream transcripts.
Enterprise tier. No training, no human review of audio.
US
OpenAI · text-embedding-3
Semantic search across your workspace's own calls.
Zero Data Retention. Embeddings never leave your tenant.
US

Known limits and how we handle them.

Every AI product has failure modes. Pretending otherwise is how trust collapses. Here are the four we watch most carefully — what can go wrong, and what we do about it.

§ 01 · Hallucination

LLMs can invent facts.

Coaching suggestions are grounded in the specific transcript — we quote the exact call moment for every insight. If the model can't cite a transcript snippet, we don't show the insight. Rep score cards show source quotes, not just judgments.

§ 02 · Bias

Models can penalize non-standard speech.

Accents, speech patterns, and cultural communication styles can skew model output. We benchmark methodology scoring across demographic groups quarterly and publish findings to Enterprise customers. Managers see raw transcripts, not just scores.

§ 03 · Over-reliance

AI shouldn't make HR decisions.

Our Acceptable Use forbids using the Service as the sole basis for firing, PIPs, or pay changes. Our product UI discourages it: leaderboards show trends, not grades; coaching is framed as suggestion, not directive.

§ 04 · Data leakage

Prompts can leak into shared context.

Every call to a model provider is isolated per workspace. No cross-tenant prompting, no retrieval across customers, no shared embeddings. Architecturally, your data cannot appear in another customer's session.

How we govern this.

Responsible AI is a system, not a slogan. Here's the operating rhythm behind what we just described.

§ 01

AI Review Board

A cross-functional group (engineering, product, legal, customer success) reviews every material change to how we use models — new providers, new prompting techniques, new automated behaviors.

§ 02

Model change notice

30-day notice to admins before adding or removing a model provider, or changing how their data is processed.

§ 03

Red-teaming

Quarterly adversarial testing by an independent firm — prompt injection, jailbreak attempts, data-leak probes.

§ 04

Bias audits

Methodology scoring benchmarked across demographic cohorts quarterly. Material disparities trigger prompt or model changes.

§ 05

Opt-out

Enterprise customers can disable any AI feature (scoring, suggestions, summarization) workspace-wide if they need to.

§ 06

Transparency

Any material change to this page is published with a changelog. Prior versions available on request.