Skip to content
Documentation

SotsAI

SotsAI is a behavioral reasoning layer designed to plug into enterprise LLM stacks
(Mistral, Azure OpenAI, on-prem, or internal models).

It helps copilots do more than answer questions —
it helps them adapt communication to real people and real situations.

What SotsAI is (and isn’t)

Section titled “What SotsAI is (and isn’t)”

SotsAI does not replace your LLM.

It acts as an interpersonal intelligence layer that your LLM can call selectively when the user’s query involves:

  • communication issues with another person
  • feedback, tension, disagreement, etc.
  • uncertainty about how a message will land

Your LLM stays in control of:

  • language generation
  • tone
  • final response

SotsAI provides:

  • behavioral insights
  • friction analysis
  • guidance on how to adapt communication

When should your LLM call SotsAI?

Section titled “When should your LLM call SotsAI?”

Use this simple rule:

If a response involves another person,
and there seems to be a communication issue,
then SotsAI should be called.

Typical examples:

  • giving or receiving feedback
  • handling a conflict
  • explaining a decision
  • handling resistance or conflict
  • persuading or aligning stakeholders
  • managing up or down

If the task is purely factual or technical, your LLM does not need SotsAI.

Why psychometric profiles matter (critical)

Section titled “Why psychometric profiles matter (critical)”

Without a user profile, SotsAI cannot:

  • reason about communication style
  • identify interpersonal friction points
  • explain why a message may trigger resistance
  • provide meaningful adaptation guidance

For this reason:

If you do not have a user psychometric profile, you should not call SotsAI.

In that case, your LLM should handle the request autonomously using its own generic reasoning.

Required

Section titled “Required”
  • context_summary (a PII-free summary of the user query)
  • user_profile (the user psychometric profile)
Section titled “Recommended (intended usage)”
  • context_summary
  • user_profile
  • interlocutor_profile (when another person is involved)

Psychometric data is therefore mandatory for SotsAI to provide value.

Where do profiles come from?

Section titled “Where do profiles come from?”

You have three options:

OptionDescription
1. Bring your own profilesUse existing DISC, MBTI, or other any other psychometric tool’s profiles if you already use them.
2. Fetch profiles via SotsAI (DISC)Use SotsAI’s DISC flow to collect and fetch profiles securely when none are available.
3. Fallback outside SotsAIIf no user profile is available, do not call SotsAI and let your LLM handle the request autonomously.

The Psychometrics section explains these options in detail.

Data & privacy model

Section titled “Data & privacy model”

SotsAI is designed for enterprise-grade privacy:

  • no raw conversations stored
  • no personal data required beyond what you explicitly provide
  • psychometric data handled securely
  • email-based invitations only when using SotsAI-managed DISC
  • encrypted in transit and at rest
  • no PII exposed in API responses

You control:

  • what data is sent
  • when SotsAI is called
  • how outputs are logged or persisted

How to get started

Section titled “How to get started”

Choose your entry point:

  • Quickstart
    Make your first API call or tool call in minutes.

  • Integration Guides
    Learn where SotsAI fits in an LLM pipeline, tool-calling patterns, retries, caching, and fallback strategies.

  • List of supported psychometric tools See the list of all supported pscyhometric tools (DISC, MBTI, etc.)

  • API Reference
    Detailed documentation for /v1/advice, optional DISC endpoints, and error codes.

  • Psychometrics
    Understand why DISC credits exist, how to invite users, and how to fetch profiles.

  • Security
    API key handling and redaction strategies.

  • FAQ & Troubleshooting
    Common integration pitfalls and how to resolve them.