Switchboard

One voice-agent engine. Every call your team keeps meaning to make — as a pluggable playbook.

● Mock mode · simulated dial CALL-E hackathon · POC
The engine — one loop, every playbook
Impact this session 0 / 3 playbooks run Lead qualified Launch unblocked Meeting saved
JV
—
—
—

Why this places across categories

Judged on impact, idea, implementation, and demo — one build answers all four.

Real-world impact
Three pains an agency actually has — each call carries an ROI: a qualified lead, an unblocked launch, a saved slot.
Quality of idea
A platform, not a bot. Use cases are configuration, not new code — most entrants ship a single narrow agent.
Technical implementation
One engine, one clean CALL-E adapter seam. The hard part is built once and reused three times.
Product & demo
The same loop does three jobs on one screen — a complete experience, not a slide.
Add a fourth use case = write a playbook, not a new app.
The adapter seam — where mock becomes real swap 1 function, keys land → live 
// Mock today, live CALL-E when the keys land —
// every playbook above runs through this line unchanged.
async function placeCall(playbook, contact) {
  return calle.calls.create({
    to:        contact.phone,
    objective: playbook.objective,
    script:    playbook.questions,
    onResult:  playbook.parseOutcome,   // → structured JSON
  });
}