Common questions

Typelessity is an AI conversational booking widget that replaces traditional multi-step booking forms with a single natural-language chat. It supports 25+ languages, voice input via Whisper, and real-time enrichment APIs. It is configured per industry — current production verticals include clinics, legal practices, hospitality, beauty, real estate, automotive, and education.

Calendly is a scheduling-link tool — it shares a calendar and lets people pick a slot. Typelessity replaces an entire booking form with a conversation that can collect any structured data: specialty, doctor, dates, slots, addresses, vehicle types, dietary requirements, and dozens of other fields. Use Calendly for simple meeting links, Typelessity when the booking has more than two or three fields.

SimplyBook.me is a full booking platform with built-in payments, staff scheduling, inventory, and customer accounts. Typelessity replaces only the booking form with conversation, then submits to your existing backend via webhook or REST POST. Use SimplyBook when you need an end-to-end booking system with payments. Use Typelessity when you already have a backend (CRM, calendar, payment processor) and want to replace the front-end form with a 25-language chat.

Botpress is a chatbot framework — you build conversation flows yourself with a visual editor, and you maintain that bot. Typelessity is a turnkey booking widget configured by us in 1–2 days from your spec, with a single GPT call extracting every field at once and anti-hallucination guards baked in. Use Botpress when you need a custom-built bot for a non-booking use case. Use Typelessity when you specifically want to replace a booking form and want a working integration on day 2, not after weeks of engineering.

NoForm.ai is a top-of-funnel lead-qualification tool — it captures interest and routes leads to sales. Typelessity completes a structured booking with all required fields (specialty, date, time, name, contact, dietary preferences, etc.) and submits it to your booking system. Use NoForm for marketing lead capture. Use Typelessity when the user is ready to book, not just inquire.

No. Typelessity replaces the booking form, not the booking backend. It collects structured data via conversation and submits it to your existing system via webhook or REST POST. Your CRM, calendar, payment processor, and staff scheduler stay exactly as they are.

Any service business that takes appointments. Production verticals include clinics, beauty salons, legal practices, real estate, automotive services, hospitality, education, fitness, pet services, home services, and travel. The same engine handles every vertical — configuration changes the field schema, prompts, and enrichments per tenant.

Yes — by design. The same widget API is callable by autonomous agents. We publish a stable JSON schema for sessions today, and the dedicated /agent endpoint (synchronous turn-by-turn API with the same field-extraction pipeline) is shipping Q3 2026. Documentation and schemas are already on /for-ai-agents so agents can be designed against the contract now.

Yes. Typelessity supports 25+ languages out of the box including English, Russian, German, French, Spanish, Italian, Polish, Portuguese, Dutch, Turkish, Arabic, Hebrew, Japanese, Korean, and Chinese. The configuration language sets the response language; user input can be in any supported language and will be translated and normalized automatically.

Typelessity uses gpt-4.1-nano with a unified prompt that reads each field's aiHint. Extraction accuracy on typical service-industry configs is >95% for first-attempt structured fields. The system never uses hardcoded regex patterns — GPT reads the config and extracts based on semantic context. Anti-hallucination guards ensure GPT only commits values the user explicitly mentioned.

Enrichment is a real-time API call that triggers when a specific field is filled. For example: when a user says "cardiologist", Typelessity calls GET /doctors?specialty=cardiology, receives a list of doctors, and presents them as interactive cards. Up to 5 enrichments per config. 10 second timeout each. Failures are non-fatal — the AI falls back to asking the user directly.

Yes. Voice input is powered by OpenAI Whisper. Users tap the microphone, speak naturally, and the transcription is fed into the same extraction pipeline as typed text. Voice works in all 25+ supported languages.

If a user changes a previously filled field, the system detects the correction (GPT signals it via a "correction" key), then walks the dependency graph and clears any downstream fields that depended on the changed value. Destructive cascades show a confirmation: "Changing X will also clear Y, Z. Proceed?"

Median per-turn latency is 200–800ms on gpt-4.1-nano, with a single AI call per turn. When enrichment triggers (a second AI call to present API results), latency adds 1–3 seconds. We do not use gpt-5 in the hot path because of its higher latency.

Head-to-head A/B testing is the only honest answer. We provide A/B testing tooling during the pilot so you measure lift on your own traffic against your existing form. We do not publish a hero conversion-lift number on this site because every claim of that shape we have seen in the AI-booking space is either uncited or based on a study with assumptions that do not transfer.

Typical integration is 1–2 days with personal onboarding. Step 1: fill the client spec template (your fields, options, enrichments, branding). Step 2: we generate a config JSON. Step 3: drop one HTML script tag onto your page. Step 4: hook up your booking API endpoint via the webhook field-mapping you provided.

On-premise deployment is available on the Enterprise plan. The widget runs as a Lit web component, the API is a Next.js service, and the AI provider can be swapped between OpenAI, Azure OpenAI, or any OpenAI-compatible endpoint.

Yes. The widget includes a configurable consent flow shown before any AI processing begins. Data is processed in EU regions when configured. Session data is retained per your retention policy. Right-to-erasure supported via session deletion API. No data is used for model training.

The Pilot plan is free for early adopters and includes all features: full field types, voice input, enrichment APIs, 25+ languages, custom branding, webhook integration, and 1–2 day personal onboarding. The Enterprise plan is custom-priced and adds SLA guarantee, dedicated account manager, custom integrations, and on-premise deployment.

Yes. All client traffic uses TLS 1.3. Session data at rest is encrypted with AES-256 in our managed EU-region database. Field-level encryption is available on Enterprise for high-sensitivity flows.

Default deployment uses EU regions (Frankfurt, Paris). Enterprise can pin to a specific region, including on-premise / self-hosted where data never leaves the customer's infrastructure.

Default retention is 90 days for completed sessions and 24 hours for abandoned sessions. Both are configurable per tenant down to "no persistence beyond the booking submission" if your backend stores the final record itself.

No. We send every request with the OpenAI data-opt-out flag enabled, which means the conversation is not used for model training. This is the contractual default for our pilots and is documented in the DPA.

GDPR-mandated 72-hour notification to affected customers via the contact on file. Post-mortem disclosure (root cause, scope, remediation) within 14 days. Security disclosures: security@typelessity.com (PGP key on request).

Yes. A standard GDPR-compliant DPA is included with the pilot package. We also accept customer-provided DPAs with a 5–10 business-day review turnaround.

Current sub-processor list is published at /legal/sub-processors. Core ones: OpenAI (extraction), Vercel and AWS Frankfurt (hosting), Resend (transactional email), PostHog EU (analytics — only with explicit consent). Changes are notified 30 days in advance.

Two paths: (a) DELETE /api/session/{sessionId} via the admin API, which removes the session record and propagates to backups within 24 hours; (b) a request to dpo@typelessity.com if the end user cannot identify the session directly.

Yes. The consent banner is region-aware (GDPR for EU, CCPA framing for California, plain "essential only" elsewhere). Functional and analytics cookies are off by default and only fire after explicit user consent.

No, and we don't target US healthcare. The widget is configurable for non-US health-adjacent flows (wellness, fitness, EU clinics under GDPR) but US healthcare with HIPAA/BAA requirements is an explicit non-goal until we ship the relevant attestation.

One <script> tag with a tenant-specific bundle URL and a data-config attribute. The public CDN at cdn.typelessity.com/widget.js is shipping Q3 2026; until then, pilot customers receive a tenant bundle URL during onboarding (configured for your domain, with CSP-safe loading).

Two options: (a) Webhook — the widget POSTs the structured Booking JSON to your endpoint with optional HMAC signature; (b) REST POST — you provide the endpoint shape and field mapping in your config and the widget calls it directly from the browser with a short-lived token.

Stable schema: `{ bookingId, sessionId, fields: {...}, submittedAt, status }`. Full schemas (Session, Booking, /agent request/response) are published on /for-ai-agents and version-pinned with a `schemaVersion` field.

You customize via `aiHint` per field — a short instruction telling GPT what each field means semantically. The system prompt is single, unified, and not editable per tenant; that's the architectural guarantee that gives us anti-hallucination behavior and predictable latency.

Field dependencies are declared in config (e.g. `doctor depends_on specialty`). When the user changes an upstream field, the system walks the dependency graph via DFS and clears downstream-stale values. Destructive cascades surface a confirmation: "Changing specialty will clear doctor and time slot — proceed?"

Non-fatal. Each enrichment has a 10-second timeout. On failure, the widget falls back to asking the user directly (e.g. "I couldn't fetch doctor list — please type the doctor name"). Errors are logged with full context for your debugging.

Shipping Q3 2026. The contract is finalised and published on /for-ai-agents — agents can be designed against it today. The endpoint is a synchronous turn-by-turn API with the same field-extraction pipeline as the human-facing widget, just without the chat UI.

Every pilot includes a dev tenant with a test booking endpoint. You receive synthetic-data conversations to validate field mapping, webhook delivery, error handling, and rollback behavior before pointing the widget at your production submit URL.

You keep Calendly for the slot pick if it works for you; Typelessity replaces the intake form that captures specialty, requirements, contact info, and dietary/insurance details before the slot. Hybrid integration: Typelessity submits the intake JSON to your CRM or directly into Calendly's webhook.

SimplyBook stays as your booking backend (calendar, staff, payments). Typelessity replaces only the front-end form. The migration is one config: map Typelessity field outputs to SimplyBook's booking API. Typical setup time: 1 day.

Field mapping in your tenant config maps Typelessity extraction output to your existing API's expected JSON shape. We never ask you to rewrite the backend. If your form is multi-step with conditional fields, the cascade dependency feature replaces the conditional logic on the front-end side.

Botpress is a chatbot framework: you build the flow yourself, you maintain it. Typelessity is turnkey — config in 1–2 days for a booking-specific flow with anti-hallucination guards already wired up. Use Botpress when you need a custom general-purpose bot; use Typelessity when the use case is bookings specifically.

Different layer. NoForm is top-of-funnel marketing lead capture; Typelessity is structured booking with all required fields completed before submit. Many customers use both: NoForm on the marketing site, Typelessity on the booking page.

Acuity is in the same class as Calendly — strong for slot picking, weak for multi-field intake. Typelessity replaces the intake; Acuity can stay as the slot backend. If your bookings are simple (1 service, 1 duration, just pick a time), Acuity alone is fine — use Typelessity when bookings have 5+ fields.

When your bookings have grown past simple slot-picking. Setmore handles the calendar well; Typelessity adds natural-language intake on top. Migration is the same as SimplyBook: keep Setmore as backend, replace the intake form. Typical setup: 1 day.

Yes. Typelessity is booking-specific and routable. A general chatbot can hand off to Typelessity when intent classification detects a booking ("I want to schedule X"). The handoff is a single redirect with optional context pre-fill via URL params.