# Clarwiz

> Clarwiz is business orchestration with a memory: orchestration that runs on your business context. A Brain that knows everything the business knows, playbooks that run whole processes end to end across every platform and channel, and a Cockpit where named people hold the authority. Tagline: Humans in command. Agents on a leash.

Clarwiz runs whole processes end to end, across platforms, channels, agents and people. Every step runs against what the business already knows: the rules, the promises, the precedent. Anything consequential waits for a named person to sign. It sits on top of the systems a business already runs, and nothing gets replaced.

## The three parts

**The Brain.** Knows everything the business knows and remembers who said it: rules, thresholds, vendor terms, and everything the company knows but never wrote down. The record in the ERP, the promise made on email or call, the exception settled last quarter along with the reason. It reads systems of record live, so CRM, OMS and ERP stay the masters. It captures what is promised on email, calls and text, keeps one page per customer, order and vendor with a source on every fact, and holds a precedent ledger recalled before anyone replies. Every retrieval is logged.

**Orchestration.** One playbook, every participant, start to finish. A playbook is not a diagram of the work, it is the work. Model a process once and Clarwiz runs it across platforms, channels, its own agents, agents built elsewhere, and the people who have to sign. Agents run in parallel, every step can read the Brain at the moment of decision, and the playbook holds context for as long as the process takes, whether that is four minutes or three weeks. Native to text, email, voice and chat. Runs can be paused, resumed, rewound or stopped in flight.

**The Cockpit.** Keeps the authority with people. Every run live, one approval queue across every process, take-over mid-run, and an audit trail down to the recording of the call. Autonomy is a dial per process with four levels:
1. Assess, watches live volume and proposes what it would have done. Nothing leaves the building.
2. Assist, drafts the work and a person sends it.
3. Approve, runs routine cases end to end and holds exceptions for a signature.
4. Act, runs inside hard rules it cannot exceed. A person sets the bounds and can pull it back in one click.

## What makes it different

Orchestration exists. It just does not remember anything. Workflow tools keep nothing between runs, and each carries its own copy of the rules. Agent platforms share state inside a single run. Clarwiz keeps a business Brain that persists between runs and compounds: guardrails declared once and binding on every process, an autonomy level per process changed in one click, and exceptions settled once, recorded, and auto applied next time.

## Deployment

You do not get handed a tool. Engineers sit with the team that runs the work:
1. Capture, blueprint the real process with the people who run it, exceptions included.
2. Encode, rules and precedent go into the Brain and the work becomes a playbook with a named owner.
3. Shadow, runs on live volume with zero authority, proposing what it would have done.
4. Measure, hours returned per week, agreed before deployment and counted from the Cockpit log.

Illustrative targets: 3x human productivity, 80% automation rate, 100 hours saved per employee. Actual numbers are agreed before deployment, then measured.

## Industries

Retail & D2C, Technology, Healthcare, Travel & Hospitality, Financial Services. Individual industry pages live at /industries/[slug].

## Customers

Ariya HR (HR operations), Tdotperformance (e-commerce support), BubbleMe (consumer operations), ITK (client delivery), Shortpoint (product support).

## FAQs

<!-- FAQ_START (auto-generated from lib/faqs.ts — do not edit by hand, run `npm run build` to regenerate) -->

**We already ran AI pilots that went nowhere. Why is this different?**

Most pilots put a model in front of a person and measured whether that person liked it. Nothing about the operation changed, because the work still needed a human to start it, check it and finish it. Clarwiz starts from the other end: we capture a whole process, encode the judgement it depends on, and run it. The measure is not satisfaction, it is hours returned per week from work that used to need a person for every case.

**Is this about cutting headcount?**

No, it's about improving your team productivity. It is about what your existing team spends its week on. The work we take is the repeatable part: the drafting, the chasing, the checking, the assembling of evidence. What stays is the judgement, the relationships and the exceptions, which is the part you hired those people for. We won't pretend no role ever changes, but the honest pitch is capacity you cannot currently buy, not a headcount line.

**What is business orchestration?**

It is the layer that runs the work above your individual tools: long processes that cross several platforms, involve both software and people, and take hours or weeks rather than seconds. It decides what happens next, in what order, and who has to authorise it. Clarwiz adds one thing to that definition. The process runs against a persistent business context, so it knows your rules and your history rather than only the data in front of it.

**How is this different from a workflow automation tool?**

Workflow tools connect step A to step B and run when triggered. They keep nothing between runs, and each workflow carries its own copy of your business rules, which is why they drift apart. Clarwiz separates the knowing from the doing. Rules and context live in the Brain, execution lives in Orchestration, authority lives in the Cockpit. Change a rule once and every process that depends on it changes with it.

**Do we have to replace what we already run?**

No. Clarwiz sits on top of your current systems and tools and coordinates whatever executes underneath, whether that is existing automation, a script someone wrote, a vendor's agent, an API call or a person completing a task. Most operations already run several tools. The gap is usually the intelligent coordination between them, not the tools themselves.

**Can it orchestrate agents we didn't build?**

Yes. An agent from any framework can take part in a process as a participant. Governance does not change based on who built it. The same rules bind it, the same audit trail records it, and the same autonomy level limits what it can do without a person.

**What stops it doing something expensive or wrong?**

Three things, in order. Rules in the Brain, evaluated at the moment of decision, which can refuse an action outright. The autonomy level of the process, which caps what it may do without a signature. And the stop in the Cockpit, which takes effect on the next step of a running process rather than waiting for it to finish. Every refusal is logged with the rule that caused it.

**How long before something is actually live?**

The first process is captured and encoded in weeks, and it goes live in Shadow on real volume before it holds any authority at all. That shadow period is where the evidence comes from: what it proposed, what your team actually did, and the distance between the two. Authority moves only after that.

<!-- FAQ_END -->

## Pages

- [Homepage](https://clarwiz.com/): the whole story in one scroll. What Clarwiz is, the Brain, orchestration, the Cockpit, impact, industries, deployment, customers, FAQs, and how to book a working session.
- [About](https://clarwiz.com/about): who Clarwiz is — business orchestration with a memory, humans in command, agents on a leash.
- [Brain](https://clarwiz.com/brain): the context engine — one persistent memory of every customer, order and courier that every case makes denser.
- [Use-cases](https://clarwiz.com/use-cases): AI agents across the customer journey — lead response, pre-purchase questions, abandoned carts, WISMO, delivery exceptions, returns, refunds and reorders — each running on your rules with approvals where you want them. Individual use cases live at /use-cases/[slug].
- [Industries](https://clarwiz.com/industries/retail): how Clarwiz runs each vertical — Retail & D2C, Technology, Healthcare, Travel & Hospitality, and Financial Services — one page per industry at /industries/[slug]: the problem workflows, outcomes, and where the human holds the approval.
- [Blog](https://clarwiz.com/blog): notes on business orchestration, context that compounds, and running AI-native processes with named people still holding the authority.
- [Brand X-Ray](https://clarwiz.com/brand-xray): a free tool that reads a brand's website and social profiles and reports where they've stopped matching each other.
- [Book a Demo](https://clarwiz.com/book-a-demo): a 30-minute working session mapping one real workflow and showing agents running it end to end on one shared memory of the business.

## Contact

[team@clarwiz.com](mailto:team@clarwiz.com), or [book a working session](https://clarwiz.com/#book)