Get started

Quickstart

Visit websiteOpen portal

This guide takes you from a fresh OrcaPulse account to a working first workflow. It is based on the features that are present in the project today: workflow steps, channel integrations, qualification prompts, CRM sync, webhook sync, AI calling, timing controls, and launch testing.

Before you start

The best Quickstart experience starts with a narrow launch goal. Before building the first workflow, decide what the flow should do, which channel or lead source it should start from, and which integrations you actually plan to use in version one.

Quickstart works best when the first launch is deliberately narrow. One live source, one qualification path, and one meaningful downstream action is usually enough to prove that OrcaPulse is configured correctly.

  • Pick the first channel: Choose the source already generating meaningful inbound demand, such as Instagram, WhatsApp, LinkedIn, or web forms.
  • Choose the first execution path: Decide whether the workflow should send WhatsApp, email, SMS, AI call, CRM sync, webhook sync, or wait steps.
  • List qualification prompts: Capture the questions or conditions that determine intent, fit, urgency, region, budget, or readiness.
  • Choose the downstream action: Decide whether the workflow should update a CRM, call a webhook, or continue into another step.
  • Check resource readiness: Confirm that the templates, providers, phone numbers, assistants, CRM connections, or webhooks needed by the first path already exist.
  • Keep scope tight: Start with one workable sequence instead of trying to cover every possible edge case on the first launch.

First launch session

The goal of the first session is to launch one working path, not to model every edge case on day one. If the first workflow can reliably capture, qualify, and trigger one correct next step, you already have a meaningful base to expand.

Launch one narrow workflow first

Start with one channel, one qualification path, and one downstream action so your team can validate behavior quickly before expanding scope.

Keep the first flow simple

Start with one clear workflow path so you can validate the step order, timing, and integrations before adding more branching.

Test every path before traffic goes live

Use realistic lead examples for high-intent, low-intent, incomplete, and escalation cases so launch-day behavior is predictable.

  • Minute 0-2: Sign in to the app and confirm you are working in the right account with the access needed to edit workflows and integrations.
  • Minute 2-4: Connect the first channel or integration you want the workflow to use, and confirm its dependencies are available.
  • Minute 4-6: Add the most important qualification prompts that capture intent and business fit.
  • Minute 6-8: Configure the first real sequence of steps, including any wait, CRM sync, webhook sync, messaging, or AI call actions you need.
  • Minute 8-10: Run test executions and confirm the workflow behaves correctly end to end.

Create the workflow

In the current product, the workflow is the main unit of execution. The best first setup is a small workflow with a clear name, a clean step order, and only the integrations you actually plan to test.

OrcaPulse supports richer execution paths than most teams need at first. Resist the urge to add every branch, retry, and handoff rule immediately. Your first goal is a workflow that is easy to inspect and easy to debug.

  • Name the workflow clearly: Use a name that reflects the lead source or execution goal so it is easy to identify later.
  • Set execution timing: Review execution delay, day and night mode, and working-hour settings if the flow should run only within certain times.
  • Enable duplicate protection if needed: Use the workflow setting that prevents duplicate leads when your launch path should avoid repeat execution.
  • Keep step dependencies explicit: If a step needs a template, provider, account, phone number, assistant, CRM, or webhook, confirm that dependency now instead of discovering it during launch.
  • Keep the first version small: A compact workflow is easier to debug than a broad automation with too many branches.

Connect channels

Channel setup is where Quickstart becomes real. OrcaPulse works best when teams begin with the channel that already creates meaningful inbound demand because that gives the fastest feedback loop on qualification quality and routing accuracy.

Connect one live channel

Choose the inbound source that already matters most to your team. A focused first launch is easier to debug and easier to teach internally.

Build one clear qualification path

Ask for intent, timeline, business fit, and the one or two details that determine whether the conversation should advance.

Push outcomes downstream

Once the workflow reaches the right step, send data to a CRM, trigger a webhook, or continue execution through the next configured action.

  • Facebook or Instagram: Best when your team needs social capture, comment or DM response, and clear source-aware qualification.
  • WhatsApp: Strong when your team needs faster response and messaging-heavy qualification with a real inbox model.
  • LinkedIn: Useful when inbound demand is more sales-driven, founder-led, or outbound-assisted.
  • Email Hub or web forms: Good when your team wants more predictable or infrastructure-backed intake and follow-up.
  • Channel tone: Match response style and qualification wording to the channel context so the conversation feels native.
  • Connection readiness: Make sure the channel or integration is actually connected before adding steps that depend on it.

Build the qualification flow

The qualification flow is the heart of Quickstart. Your first flow should be short, useful, and easy for operators to understand. Do not try to model every edge case in version one.

The fastest way to make Quickstart hard is to ask too many questions or to collect information that never changes the workflow outcome. The first flow should gather only the details that actually affect routing, urgency, handoff, or follow-up.

  • Ask only decision-making questions: Focus on intent, company type, budget, location, timeline, or use case depending on your business.
  • Keep the conversation natural: Phrase qualification prompts like real follow-up messages rather than a rigid survey.
  • Define success thresholds: Decide what counts as qualified, unqualified, incomplete, or review-required before launch.
  • Map prompts to steps: Make sure each question has a clear follow-up action in the workflow.
  • Leave room for human review: If some replies should not be auto-routed, define a safe review or handoff path instead of over-automating the first release.
  • Document the reasoning: Keep the first flow understandable so it is easy to edit when test results come back.

Route and notify

Quickstart is not complete until OrcaPulse creates a real execution outcome. In the current codebase, that usually means moving from one workflow step to the next and, when needed, sending data into a CRM or webhook integration.

If the first workflow does not create a concrete outcome, the test is incomplete. You should be able to point to what changed after qualification: a record synced, a message sent, a call started, a webhook fired, or a lead moved into the next step.

  • CRM sync: Push lead data into a configured CRM connection when the workflow reaches that step.
  • Webhook sync: Send payloads to an external webhook when the flow should notify another system.
  • WhatsApp, SMS, or email steps: Use messaging actions when the launch path should continue with outbound follow-up.
  • AI call: Use calling steps when your workflow needs voice-based follow-up and the required calling configuration is ready.
  • Wait steps: Add delays between actions when execution timing matters.

Test before launch

Every Quickstart should end with deliberate testing. Do not assume the workflow is correct just because the happy path works once.

Real launch quality comes from testing multiple behaviors, not one successful run. OrcaPulse workflows often depend on timing, prompts, routing criteria, connectors, and resource state, so each of those assumptions should be challenged before traffic goes live.

  • High-intent test: Confirm that a strong lead moves through the expected step sequence from start to finish.
  • Low-intent test: Verify that weaker demand does not trigger the same follow-up path as stronger leads.
  • Missing-information test: Check how OrcaPulse handles vague replies, incomplete data, or abandoned conversations.
  • Step execution test: Make sure waits, messages, CRM sync, webhook sync, and AI call steps all execute as configured.
  • Delivery test: Confirm CRM updates, webhook events, messages, or calls arrive where expected.

Team onboarding

A technically correct launch can still fail if the people operating the workflow do not understand how it behaves. Before turning on real traffic, align the people who will edit, test, and monitor the flow.

  • Show the workflow structure: Walk operators through the actual step order and what each step is supposed to do.
  • Explain test expectations: Make sure the team knows how to validate CRM sync, webhook sync, messaging, and AI call behavior.
  • Clarify who edits the flow: Decide who is responsible for updating steps if the first launch reveals problems.
  • Create a feedback loop: Ask testers to log confusing prompts, weak step sequences, and edge cases so the workflow can improve quickly.
  • Review launch metrics: Track execution success, failed actions, and whether the live path is doing what the workflow was designed to do.

Next steps

Once Quickstart is live, the next phase is to harden the workflow. That usually means refining the step sequence, improving qualification prompts, adding more integrations, and testing additional real execution cases.

If you are ready to continue building, the most useful next docs pages are usually Workflow Setup, Qualification Flows, Routing Rules, and Automation Connectors.

If you want to continue inside the product, go deeper into configuration inside app.orcapulse.ai. If you need the high-level architecture again, revisit Platform Overview.