Skip to content

Fetching psychometric profiles

Fetching psychometric profiles is a support step, not a core feature on its own.

Its only purpose is to ensure that psychometric data is available when calling /v1/advice.

If profiles are missing, SotsAI must not be called. In that case, your system should rely on your LLM alone.


You should fetch a profile before calling SotsAI when:

  • your LLM needs to reason about a specific person
  • you do not already have a psychometric profile in memory
  • profiles were collected via SotsAI DISC invitations

Typical cases:

  • first interaction with a user
  • cold start after cache eviction
  • asynchronous profile completion after invitation

If you already have a valid profile in memory or cache, do not fetch again.


When you fetch a DISC profile via SotsAI:

  • the profile is resolved using:
    • your organization identifier
    • the user’s email (used only as a deterministic lookup key)
  • the identifier is non-reversible
  • no raw psychometric results are stored or exposed

SotsAI acts as a secure proxy:

  • profiles are fetched on demand
  • results are normalized into the profile format expected by the advice engine
  • nothing is persisted by default

A common pattern looks like this:

1. Your system needs to generate advice involving a specific person
2. You check whether a user psychometric profile is already available
3. If not:
- attempt to fetch the profile via SotsAI
4. If the profile exists:
- include it when calling `/v1/advice`
5. If the profile does not exist:
- do not call SotsAI
- optionally trigger a DISC invitation
- handle the current request without behavioral reasoning

This logic should live in your orchestration layer, not in prompts.


A missing profile usually means the user:

  • has not been invited yet, or
  • has not completed the DISC assessment

In that case, you must not call SotsAI.

You have two valid options:

  1. Gate the feature
    Inform the user that personalization requires a psychometric profile and offer a way to complete the assessment.

  2. Handle the request without SotsAI
    Let your LLM generate a generic response without behavioral reasoning.

Optionally, you may trigger a DISC invitation so that future requests can benefit from SotsAI once the profile exists.


Fetching profiles via SotsAI is designed for enterprise use:

  • no raw psychometric reports are exposed
  • no personal data stored in clear
  • identifiers are deterministic but non-reversible
  • data is encrypted in transit
  • profiles are only returned to the authenticated organization

SotsAI never infers or guesses profiles.