QuickFlo – I got sick of exporting workflows and everything that came with them

zajasa1 pts1 comments

Build the workflow once. Ship it as a product. | QuickFlo

Workflow infrastructure<br>Build the workflow once. Ship it as a product.<br>Describe what you need and get a workflow that runs. Package it with its triggers, data stores, and dashboards, and install it across your customer base.

@your-org/contact-centerv1.0.0<br>Workflowssync-agents · publish-roster4<br>Triggerson-schedule · on-webhook3<br>Data Storesagents · skills · settings3<br>Dashboardscoverage-by-skill1<br>AI Toolsreassign_skill2<br>Connectionscrm · slack · dialer3<br>Everything in one package.ready to install

Describe the problem. Watch the solution build in real time.

Build it now

↵ to build.Free to start, no card needed. Your workflow starts building the moment you're in.

The idea<br>A workflow is not a solution.<br>It is one piece. A solution also needs the data it reads, the dashboards over it, the triggers that fire it, and the connections wired per customer. On most platforms each of those is a separate thing you rebuild by hand for the next customer. We make every piece shippable, so the whole solution travels in one package.

01 i WorkflowsReal programs on a scalable engine: traced, auditable, and editable.workflow: dedupe-leads<br>on-csv-uploaded<br>normalize<br>dedupe<br>upsert.contacts<br>notify.slack

3 workflows · 2 sub-workflows · opens in builder

02 i TriggersThe events that start a workflow: webhooks, schedules, chat, and platform events. Live the moment you save.2 triggers · live endpoints<br>on-csv-uploaded<br>POSTrun.quickflo.app/w/@acme/csv-uploadcopy

ai-agent-chat<br>Ask the agent to schedule a follow-up…↑

03 i Data storesQueryable tables the workflows read and write, with dashboards and BI built in. No external warehouse.data store: settings · 5 rows<br>default_timezone"America/New_York"<br>default_retry_count3<br>ai_routing_enabledtrue<br>default_disposition"follow-up"<br>dialer_campaign_id"Q2-OUTBOUND"

flexible key/value tables for config, data delivery, dashboards, caching, and anything else your workflows need to remember

04 i DashboardsCharts and metrics read straight from those tables. They ship inside the solution without needing an external BI platform.dashboard: agent-performance<br>CALLS / HR<br>14.2

TALK TIME<br>4:08

CONVERT<br>23%

any data-store's data can be shown in a dashboard automatically, no external BI platform required

05 i AI toolsAny workflow can be exposed as a tool an agent can call. The same program runs, traced, with approvals on writes.tool: tools.ai-scheduling-agent<br>descriptionA workflow exposed as a function. Callable from any agent.<br>requires approvaltrue<br>parametersuser_request: string

✓agent called tools.ai-scheduling-agent<br>✓arguments and result logged to the run<br>●held for approval before it ran

Every tool an agent calls is one of your workflows. Fully traced, with approval gates on writes.

06 i Connections and mappingsPer-customer wiring: credentials, secrets, and field maps supplied at install. The logic stays the same.install-time bindings · 3 mappings<br>connections.salesforce→ salesforce_prod<br>env.SLACK_WEBHOOK_URL→ ••••••••<br>env.DEFAULT_TIMEZONE→ America/New_York

customer-supplied at install time

How it is built<br>From a blank canvas to an installed solution.<br>The same surfaces, step by step: design the workflow, wire the triggers, model the data, read it back on a dashboard, then package it and install it for the next customer.

01Build02Trigger03Run04Data store05Dashboard06Package07Install

Drag steps onto the canvas to build you business logic, or build with the AI builder.

The model<br>Stop rebuilding it for every client.<br>Today every client is a fresh project: a new build, somewhere to host it, one more thing to maintain. Package it once and that ends. Install the same solution for every client you serve, each on their own data and settings, and let us run the hosting, the scaling, and the uptime. You ship in days, not quarters, and the work you did for the last client compounds into the next.

@your-org/contact-centerinstalled by 5 customers<br>latest v1.4.0<br>5installs<br>128kruns / mo<br>1needs attention

acme-cov1.4.0ok<br>globexv1.4.0ok<br>initechv1.4.0ok<br>hooliv1.4.0ok<br>pied-piperv1.3.2sync failing<br>one publish updates every install

“QuickFlo lets us say yes to the custom tooling our clients ask for. Schedule editors, contact database scrubbers, skill rotators: shipped in weeks, hosted forever.”

Founder, Square O

Integrations<br>Connect to the systems you already run on.<br>200+ step types plus first-class connections to the tools your operation depends on. If it has an API, you connect to it.<br>Don't see your system? Let's talk . We can probably connect to it.

HubSpotSlackSalesforceTwilioZoom CXWebex CCGenesysEmailSMSRESTAny API

Solutions<br>This morning's report, in Slack. The contact list, scrubbed overnight. The new agent, provisioned before their first coffee.<br>Pick a fix. We've already built it.

list-sync.q1-outbound

SalesforceDialer<br>247<br>Synced

12<br>DNC blocked

2.5s<br>Avg sync

List Sync<br>Leads from CRM to dialer in seconds. Enrich, modify, or apply your DNC list to...

workflow data agent build install solution

Related Articles