Where Zara fits

You say what you need. She builds it in front of you.

No template library to search through. No empty grid waiting for you to know what an agent architecture is supposed to look like. You type the job, and Zara starts filling seats.

She tells you why, not just what

Every pick comes with a reason, written out where you can see it, not buried in a changelog you'll never open.

No seat is a single point of failure

Every role on the panel comes with at least one alternate, already vetted, already labelled with what makes them different.

Open any seat, and the configuration is real

Not a black box with a name attached. An Operating Brief in plain text, an engine choice with the dials exposed, and the domain parameters that make the difference between a generic underwriter and one that actually knows UAE rules from Saudi ones.

The shape of the team, before you commit to it

A dependency graph most platforms only show you after something's broken. Zara shows it to you before you press deploy.

You can trace exactly what happens between the application landing and the decision leaving. Not because someone documented it after the fact, but because that's what the map already shows, before a single case has run through it.

The conversation is with the process owner, not the developer

Common questions about Zara

SPAR isn't a slide. It's the canvas itself

Every workflow you build runs the same four stages, enforced by the runtime, not a policy someone can quietly skip under deadline pressure.

SSource: deterministic intake, hashed at ingest

PProcess: specialists reason in parallel

AAdversarial: a different model tries to break it

RRelease: a named human signs off

The canvas above runs Consolidation mode: three specialists work different parts of the same marine claim, merge into one record at the Chief Claims Officer node, and only then face the Adversarial Agent. Fastest path, lowest token cost, right for the volume of claims that don't need three independent opinions.

For the calls that do need three opinions

Consensus mode runs the same canvas differently: specialists answer independently, and where they land, together or apart, is itself the signal worth reading.

Consolidation and Consensus aren't two products. They're the same runtime, the same guardrails, the same Adversarial stage, applied at different intensities depending on what the case in front of you is actually worth getting wrong.

Every node opens to real configuration

Not a prompt box with a friendly name. A full inspector: system prompt, temperature, token limits, and the guardrails that apply before anything downstream ever sees the output.

On by default

PII check, schema strict, require verification

Guardrails a node ships with, visible and toggleable per agent, not hidden three settings menus deep.

Named limits, not hopes

Reasoning steps, tool calls, allowed tools

An agent's ceiling is a number someone chose on purpose. A deterministic fallback covers what happens if it's ever reached.

The Stable: every agent this team can draw from

When Zara says "Dr. Neha is in," this is where Dr. Neha actually lives. 168 named agents, all mapping to six universal archetypes. Only the domain knowledge inside changes, which is why a new solution takes weeks, not months.

Deployment: where this actually runs

Same harness, wherever the data has to stay. Model choice is configuration, per agent, at runtime, not a decision you make once and live with.

Versioned, published, and still yours to reproduce

A run from March and a run from September should agree with each other if nothing about the case actually changed. That's not a promise you take on faith, it's what pinning the version, the prompts, and the rule pack together is built to guarantee.

Common questions about Zen Studio

Every deployed workflow, at a glance

Filter by specialty. Check status, version, and when it last changed, without opening a single run.

Every case, tracked while it's actually happening

Click into a workflow and the individual pipeline runs underneath it are right there, named, timed, and stepped through in real time, not reconstructed after the fact.

Every decision, explained in language a person can actually read

Click any node in a completed run. The rationale is a sentence, not a log a specialist has to translate for the compliance team.

The control plane: always on, owned by nothing

The parts of the platform that stay separate from any single workflow, because governance a workflow could switch off isn't governance.

Silent Agents: work that starts before anyone has to ask

Proactive triggers that sit above your systems of record. Event or threshold based, and governed by the same SPAR cycle as human-initiated work.

Common questions about Zen Pilot