AI receptionist for Austin locksmith shops — plugs into your stack
Austin locksmiths run more tech than most. It's the nature of the city — you already have Workiz or Jobber, a Google Calendar with 6 different tech feeds, Stripe for deposits, a Twilio number or two, and probably a Zapier flow you set up in 2023 and haven't touched since. We built TheKeyBot's AI receptionist to slot into that stack, not replace it on day one.
inbound call
│
▼
┌─────────────────────┐
│ TheKeyBot AI │ ← handles voice, quote, booking, deposit
└──────────┬──────────┘
│
webhook: job.booked
│
┌───────┴───────┐
▼ ▼
Workiz Google Calendar
(existing) (tech dispatch)
│
└──► Stripe (deposit captured)
│
▼
SMS → customer
SMS → on-call techThe architecture: TheKeyBot is a real-time voice AI + CRM + dispatch stack that sits in front of your Twilio/ATT number, drops booked jobs into Workiz or Jobber via HMAC-signed webhook, pushes deposits to Stripe, fires a Google Calendar event for the tech, and optionally rolls call-source attribution back to Google Ads. API-first, EN/ES native, sub-400ms turn latency, SOC 2 in progress. $500/month flat, no per-seat fees, no long-term contract.
Integrations for Austin shops
Everything an established Austin locksmith shop already uses, already supported.
Booked jobs sync as new work orders; deposits flow to Stripe.
Client + quote objects mapped; recurring jobs supported.
Job push only; deposit handled in TheKeyBot.
Per-tech calendar, automatic blackout-window respect.
In-call deposit links; automatic receipt; refund flow.
Any downstream tool (Airtable, Slack, Notion, Twilio SMS).
Port existing Austin number or forward from your current provider.
Source-attributed lead data pushed back to GAds via conversion API.
Beta — rolling to Austin shops Q2 2026.
Technical capabilities
Voice layer
- • Low-latency streaming speech-to-text + text-to-speech
- • ~200-400ms response latency per turn
- • Barge-in support (caller can interrupt AI)
- • Noise-robust (highway, shop, parking-lot environments)
- • Native EN + ES; auto language detection per turn
Reasoning + pricing
- • Year/make/model key-type lookup (40k+ entries, updated quarterly)
- • Mileage-band pricing engine (circle or polygon service areas)
- • Configurable hard-stops (out-of-area, illegal vehicle types, etc.)
- • Escalation rules (owner cell, specific tech, callback scheduling)
Observability
- • Full call transcripts with timestamp + speaker diarization
- • Per-call quality score (6 dimensions, explainable)
- • CSV export; nightly S3 drop for shops with BI tools
- • Webhook delivery log with replay
Security + compliance
- • PCI-SAQ A for payment flow (Stripe tokenization, no card data stored)
- • Two-party consent greeting for TX compliance
- • SOC 2 Type II (in progress, Q3 2026)
- • HMAC-signed webhooks
- • TLS 1.3 across all endpoints
Austin operational specifics
Austin's locksmith market is smaller than Houston's but technically more demanding. A few implementation notes we've learned from deploying to 30+ ATX shops.
SXSW / ACL / F1 surge scripts
Event-window configs are standard in Austin. You define the date range, premium mileage rate, and any capacity cap (e.g., "do not accept new jobs inside the Zilker zone after 10 PM during ACL weekend"), and the AI enforces it without you logging in.
West Austin geographic corrections
The AI's distance engine knows that a call from Bee Cave is not a short drive, even though Google Maps shows 12 miles — because 71 and 360 at 5 PM add 25 minutes. Mileage pricing factors in typical Austin drive-time, not just distance.
Tesla + EV key programming volume
Austin has disproportionately high Tesla, Rivian, and other EV key-card call volume. TheKeyBot's pricing database includes EV-specific entries (NFC key cards, phone-as-key provisioning, dealer-only vehicles flagged for referral). This matters in almost no other TX metro.
Call-source attribution for Google Ads
Austin locksmith shops run heavy paid-search spend. TheKeyBot tags inbound calls by source (GAds campaign ID, organic, direct, referral) and pushes conversion events back to Google Ads via offline conversion upload — letting you optimize spend on actual booked jobs, not just clicks.