Use cases
Recruiting & Onboarding

Candidate dossier for the next interview loop

Everything the team has ever said about this candidate, surfaced in one place.

Best for
Pre-loop sync
Primitives
  • Actors
  • Interactions
  • Stance
Side by side

Token Usage, with and without Sentra

~48k tokens saved60% less work for the agent

Without Sentra
~80k tokens
With Sentra
~32k tokens
Claude (without Sentra)
  1. Pull the candidate's record from the ATS. Parse role, source, recruiter, prior loops.
  2. Search Gmail for messages with the candidate's address. Skim every thread.
  3. Search Slack DMs and channels for mentions of the candidate's name. Filter to material posts.
  4. Pull calendar events the candidate has attended. Read meeting notes.
  5. For each prior interviewer, search their feedback artifact — scorecard, notes, DM to recruiter.
  6. Manually extract per-interviewer stance, concern, and evaluated dimension.
  7. Compose the dossier.
~80k tokens
Claude + Sentra
  1. Resolve the candidate as an Actor — joined across ATS, calendar, email, and Slack mentions.
  2. Pull every internal Interaction discussing them, ordered by date.
  3. For each prior interviewer, typed stance, concern, and evaluated dimension are extracted.
  4. Pull source and recruiter context from the candidate's Actor relations.
  5. Compose the dossier.
~32k tokens
Agent prompt
You are preparing the interview loop for {{candidate_name}}.

Using Sentra:

1. Resolve {{candidate_name}} as an Actor — match across ATS, calendar invites, email threads, prior Slack mentions, and any referral source.
2. Pull every internal Interaction where the candidate has been discussed, ordered by date.
3. For each prior interviewer, extract their stance, their concern (if any), and the dimension they evaluated.
4. Pull external-source context: referral notes, recruiter intake, prior loop notes if any.

Output:

- Snapshot card: name, role, level under consideration, source, recruiter, hiring manager.
- Interviewer map: who they've already met, the dimension covered, the stance.
- "What's still open" — the dimensions not yet evaluated, mapped to the panel members who will cover them.
- Three suggested probing questions per remaining interviewer based on prior concerns.

Sentralize your company.

Remember what matters.

Preferences

Subprocessors include Amazon Web Services, GitHub, Slack, Google Cloud Platform, and OpenAI.

© 2026 Dynamis Labs Inc. All rights reserved.