Replace Goodcall AI Receptionist With a Vertical Voice Agent

TL;DR — Goodcall's $79/$129/$249 plans cover the basics but charge $0.50 per extra unique caller and have shallow PMS/CRM writes. A vertical agent (HVAC, salon, dental — pick one) on Twilio + OpenAI Realtime gives you direct DB writes and predictable pricing.

What you'll build

A vertical AI receptionist for service businesses (e.g. HVAC) that answers, qualifies, books into your real schedule (ServiceTitan/Housecall Pro/Jobber), takes deposits, and SMSes confirmations — replacing Goodcall's "skills" with real Python tools.

Prerequisites

1. A clear vertical pick (HVAC, salon, dental, real estate).
2. A real PMS/CRM API: ServiceTitan, Housecall Pro, Jobber, Square Appointments, etc.
3. Twilio number with SMS enabled.
4. OpenAI Realtime + Python 3.11.
5. A list of "skills" Goodcall handles for you today (these become tools).

Architecture

flowchart TB
  C[Caller] --> TW[Twilio]
  TW --> AGENT[Vertical Agent]
  AGENT --> PMS[PMS/CRM]
  AGENT --> SMS[Twilio SMS]
  AGENT --> PG[(Postgres logs)]

Step 1 — Define vertical-specific tools

```python @function_tool async def find_open_slots(zip_code: str, day: str) -> list[dict]: techs = await pms.list_techs_by_zip(zip_code) slots = await pms.aggregate_slots(techs, day) return slots[:5]

@function_tool async def book_job(name: str, phone: str, address: str, slot_id: str, issue: str, urgency: str) -> dict: job = await pms.create_job(name=name, phone=phone, address=address, slot_id=slot_id, summary=issue, urgency=urgency) await sms.send(phone, f"Confirmed: tech arriving {job.eta_window}. Ref {job.ref}") return {"ref": job.ref, "eta": job.eta_window} ```

Step 2 — Vertical prompt

```md You are the dispatcher for ABC Heating & Cooling. Greet warmly, collect: caller name, callback number, service address, issue (no heat / no cool / no hot water / other), urgency. Use find_open_slots to read live capacity, never invent a time. Confirm slot back, then call book_job. SMS confirmation is automatic. Hours: 7am-7pm M-Sat. Emergency calls always escalate. ```

Step 3 — Twilio voice handler

Standard Media Streams bridge — same as the Synthflow tutorial.

Step 4 — Vertical knowledge

Embed your top 30 FAQs (warranty terms, brands serviced, service area zip codes) as a single get_faq tool that does a simple keyword match against a JSON file:

```python @function_tool async def get_faq(query: str) -> str: return await faqs.search(query) ```

Step 5 — Pricing logic that Goodcall can't easily do

```python @function_tool async def quote_diagnostic(zip_code: str) -> dict: pricing = await pms.get_zone_pricing(zip_code) return {"diagnostic_fee": pricing.diag_fee, "after_hours_fee": pricing.after_hours_fee, "applies_to_repair": True} ```

Step 6 — Cutover from Goodcall

Forward your business number to the new system; keep Goodcall live as fallback for one week. Compare booked-job rate.

Step 7 — Reporting

Goodcall gave you a dashboard. Replace with a 60-line Next.js page reading from call_sessions (calls/day, booking rate, top intents).

Common pitfalls

• PMS write rate limits. ServiceTitan caps writes/min; queue with retry.
• Prompt drift across techs. Lock the prompt in CI; do not let dispatchers edit it.
• SMS opt-in. Always confirm SMS consent before sending.

How CallSphere does this in production

Explore a live demo and compare current plans to find the right fit for your business.

FAQ

What if Goodcall's "skills" library is enough? Stay on Goodcall — this guide is for teams hitting the wall.

Cost per call? ~$0.04/min OpenAI + Twilio = ~$0.07/min total.

HIPAA? For dental/medical, sign BAAs and add audit logs.

Multi-location? Pass location_id into every tool; one agent serves N locations.

What about Goodcall's $0.50 per unique caller? Self-host = no per-caller fee.

Sources

• Goodcall pricing
• ServiceTitan AI voice
• OpenAI Realtime
• /pricing

Replace Goodcall AI Receptionist With a Vertical Voice Agent: production view

Replace Goodcall AI Receptionist With a Vertical Voice Agent is also a cost-per-conversation problem hiding in plain sight. Once you instrument tokens-in, tokens-out, tool calls, ASR seconds, and TTS seconds against booked-revenue per call, the right tradeoff between Realtime API and an async ASR + LLM + TTS pipeline becomes obvious — and it's almost never the same answer for healthcare as it is for salons.

Shipping the agent to production

Production AI agents live or die on three loops: evals, retries, and handoff state. CallSphere runs 37 agents across 6 verticals, each with its own eval suite — synthetic call transcripts replayed nightly with assertion checks on extracted entities (date, time, party size, insurance, address). Without that loop, prompt regressions ship silently and you only find out when bookings drop.

Structured tools beat free-form text every time. Our 90+ function tools all enforce JSON schemas validated server-side; if the model hallucinates an integer where a string is required, we retry with a corrective system message before falling back to a deterministic path. For long-running flows, we treat agent handoffs as a state machine — booking → confirmation → SMS — so context survives turn boundaries.

The Realtime API vs. async decision usually comes down to "is the user holding the phone right now?" If yes, Realtime; if no (callback queue, after-hours voicemail), async wins on cost-per-conversation, which we track per agent in 115+ database tables spanning all 6 verticals.

FAQ

How does this apply to a CallSphere pilot specifically? Setup runs 24 hours, the trial is 7 days with no credit card, and pricing tiers are $19, $49, $99, and $149 — so a vertical-specific pilot is a same-week decision, not a quarterly project. For a topic like "Replace Goodcall AI Receptionist With a Vertical Voice Agent", that means you're not starting from scratch — you're configuring an agent template that's already been hardened across thousands of conversations.

What does the typical first-week implementation look like? Day one is integration mapping (scheduler, CRM, messaging) and prompt tuning against your top 20 real call transcripts. Day two through five is shadow-mode running, where the agent transcribes and recommends but a human still answers, so you can compare side-by-side. Go-live is the moment your eval pass-rate clears your internal bar.

Where does this break down at scale? The honest answer: it scales until your tool catalog gets stale. The agent is only as good as the integrations it can actually call, so the operational discipline is keeping schemas, webhooks, and fallback paths green. The platform handles the rest — observability, retries, multi-region routing — without your team owning the GPU layer.

Talk to us

Explore a live demo and compare current plans to find the right fit for your business.