Offbrain Platform · Public Trust

Turn a loose research need into a scoped demo workspace.

Spark takes a sentence, a brief, or pasted notes and returns the apps, audience path, fieldwork plan, and proof needed for a client-ready demo workspace. The public page stays high level—a buyer sees capabilities first, then chooses whether to request saved access.

Scope a demo workspace See the respondent path
$100–$500 typical paid sessions · 7-day recruit · NDA-aware
Looper Agent
Hi! I'm Looper. I can help you find the right path whether you're here to participate in research or run it.
01.platform oaths

Our Non-Negotiable Commitments

Offbrain is built on cryptographic proof and absolute guarantees. We write these promises directly into our platform governance engine.

01

Guaranteed Creator Pay

We enforce transparent, upfront payout rates (typically $100–$500 per session). All earnings are automatically disbursed to the creator's account within 48 hours.

02

AI Ingestion Exclusion

No creator likeness, session video, audio recordings, or workspace evidence files will ever be ingested into foundation model training. Your data remains yours.

03

NDA-Protected Sandbox Sessions

All fieldwork run on Pulse is governed by automated, cryptographically signed NDAs. Workspace data access is strictly gated by role-based workspace permissions.

02.the loop

One rough ask becomes a demo workspace — four moves, in order.

01capture

Start with the rough research need.

A sentence, brief, or pasted notes. Spark drafts the first study shape and names its assumptions out loud.

module · Spark
02respondents

Show paid terms before anyone commits.

Creators are the first respondent example: pay, duration, consent, schedule, and data use stay visible from the invite onward.

module · GET
03fieldwork

Use the fieldwork shape the study needs.

Sessions, async tasks, or review work — without forcing a single process. State and recordings stay recoverable.

module · Pulse
04decisions

Return evidence, assumptions, and next actions.

Transcripts, clips, themes, and a client-ready readout. The demo makes the output legible before access is granted.

module · Core
03.spark work log
visible by default

Spark shows its work in plain steps.

Every Spark run produces a public ledger: what was captured, what was assumed, what was cited, what was invoked, what was decided. Private model reasoning stays private — the audit surface does not.

Try a Spark scope
Unsaved until signup. The public demo never auto-saves payloads.
spark · session 0001 · unsavedDraft
01
Capture
Waiting for one research goal and verified work context.
Awaiting
02
Assumptions
Ready to fill once Spark receives the rough ask.
Ready
03
Evidence
Sources, citations, and contradiction surfaces.
Ready
04
Tool steps
Which apps are invoked, in what order, with what input.
Ready
05
Decisions
Open questions, next actions, and review checkpoints.
Pending
04.modules

Focused apps, one Spark assistant.

01
GETRecruit · public-facing
Find the right people. Show the terms.Targeted recruitment with pay, consent, schedule, and data-use surfaced up front. Creators are the first audience; the same shape extends to B2B panels.
statusLive · 542 creators
latencySearch · 696ms
phasev3
02
GrabSource · specialist
Find public source signal.Internal source intake and corpus preparation. Public-source discovery, dedup, and consent tagging before anything enters a study.
statusBeta
sources12
phasev2
03
PulseFieldwork · researcher
Run the fieldwork the study needs.Sessions, recordings, async tasks, and review work. Live state, recoverable transcripts, calendar and Daily.co rooms.
statusLive
sessions37 running
phasev3
04
CoreEvidence · researcher
Turn fieldwork into a readout.Transcripts, clips, notes, and themes assembled into client-ready evidence. Citations stay attached to the source moment.
statusLive
studies8 in review
phasev3
05
InjectModules · specialist
Add specialist tools when the work needs them.Domain modules that attach to the platform: briefs, screeners, guides, study settings, and category-specific synthesis.
statusActive
modules6
phasev2
06
AdminGovernance · super admin
Govern access. Audit everything.Super admin controls plans and overrides; admins manage client and project access. Every action lives in an audit log.
statusLive
audit logUp to date
phasev3
05.access

Global access stays with super-admin authority.

Public surface

Creator

Can

Invites, consent, schedule, session receipts, payout history.

Cannot

No client workspace access.

Buyer · subscribed

Client

Can

Scope studies via Spark, see subscribed projects, review app outputs.

Cannot

No global access changes.

Workspace owner

Admin

Can

Manage client and project access; assign modules within active subscription plan.

Cannot

No plan or app global authority.

Final authority

Super admin

Can

All plans, apps, overrides, and audit logs. Every action is cryptographically signed.

Cannot

Nothing.

06.public entry

Clients scope work. Respondents see termsbefore they commit.

For clients

Scope the demo workspace first.

Start with a rough research goal. Spark returns the app mix, audience path, evidence shape, and review points. Nothing is saved until you sign up and are approved.

01Goal — paste a sentence or brief
02Scope — Spark drafts the study shape
03Modules — see which apps are invoked
04Access — request saved workspace
For respondents

Participate only after terms are clear.

Creators are the first use case. Pay, duration, consent, schedule, data use, and not used for AI training stay visible before any commitment.

01Offer — amount, duration, topic, category
02Consent — data use, retention, and disclosure
03Schedule — pre-selected with three alternates
04Receipt — paid in 48 hours · shareable

Request saved demo workspace access.

Spark payloads stay unsaved until signup and approval. Lite modules mimic the real apps without fake live handoffs. The respondent path remains separate from client workspace access.