Developer reference

SLAtech v1 API — public endpoints

9 public endpoints — bot query (sync + streaming), lead capture, landing-page payloads, pricing JSON, public stats, widget analytics. Curl examples are copy-paste runnable; auth notes and rate-limit headers are documented in the FAQ at the bottom.

Endpoints

9 routes, plain JSON over HTTPS

POST /v1/query/ask

Submit a visitor question and receive a RAG-grounded answer + citation sources.

curl -X POST https://api.slatech.ai/v1/query/ask \
  -H 'Content-Type: application/json' \
  -H 'X-Site-Id: {SiteId GUID}' \
  -d '{
        "question": "Do you accept walk-in appointments?",
        "topK": 10,
        "scoreThreshold": 0.5
      }'
POST /v1/query/ask-stream

Same as /ask but streams the answer token-by-token via Server-Sent Events.

curl -N -X POST https://api.slatech.ai/v1/query/ask-stream \
  -H 'Content-Type: application/json' \
  -H 'X-Site-Id: {SiteId GUID}' \
  -d '{ "question": "How do you handle no-shows?" }'
# SSE events: { event: "token", data: "..." }, { event: "done", data: { sources: [...] } }
POST /v1/leads

Capture a visitor's contact details and intent into the tenant's CRM webhook.

curl -X POST https://api.slatech.ai/v1/leads \
  -H 'Content-Type: application/json' \
  -H 'X-Site-Id: {SiteId GUID}' \
  -d '{
        "name": "Jane Doe",
        "email": "jane@example.com",
        "message": "Interested in dental whitening pricing"
      }'
GET /v1/landing/{site}/{lang}/{slug}

Fetch a published landing-page payload (FAQ + features + stats) for SEO renderers.

curl https://api.slatech.ai/v1/landing/beauty/en/hair-salons
GET /v1/landing/{site}/sitemap-slugs

List every published (slug, lang, type) tuple for a vertical — feeds dynamic sitemaps.

curl https://api.slatech.ai/v1/landing/med/sitemap-slugs
GET /v1/pricing/{site}/{lang}

Return the live pricing tiers (in tenant currency) for a vertical + language.

curl https://api.slatech.ai/v1/pricing/general/en
GET /v1/public-stats/{siteId}

Public per-site stats — total questions, resolution rate, leads collected, active-since days.

curl https://api.slatech.ai/v1/public-stats/{SiteId GUID}
POST /v1/query/ask-multimodal

Stub returning 501. Reserved for image-upload + question (Q4 2026 — see /en/roadmap/). SDKs probe this endpoint to detect support and fall back to /v1/query/ask.

curl -X POST https://api.slatech.ai/v1/query/ask-multimodal \
  -H 'Content-Type: multipart/form-data' \
  -F 'question=Is this rash an allergic reaction?' \
  -F 'image=@/path/to/photo.jpg'
# Returns 501 today; expected shape lives in the response body.
POST /v1/widget/track

Anonymous visitor analytics — page view, intent classification, dwell time.

curl -X POST https://api.slatech.ai/v1/widget/track \
  -H 'Content-Type: application/json' \
  -d '{ "siteId": "...", "event": "page_view", "url": "https://example.com" }'
Developer FAQ

Auth, rate limits, SDKs, webhooks

How do I authenticate against the SLAtech API?
Most v1 endpoints accept a Site-Id header (X-Site-Id: {GUID}) that scopes the request to a single tenant's data. Pro-tier and above customers receive an additional bearer token for server-to-server calls; contact info@slatech.ai to provision one. All requests must be HTTPS — plaintext HTTP is rejected at the edge.
What's the rate limit?
Per-site limits: Starter 60 RPM, Pro 600 RPM, Enterprise 6000 RPM. Headers X-RateLimit-Remaining + X-RateLimit-Reset accompany every response. Bursts up to 2× the limit are tolerated for short windows; sustained over-limit returns 429.
Is there an OpenAPI spec?
An OpenAPI 3.1 spec ships in the Enterprise tier and is on the public roadmap for Pro. Until then, this page is the canonical reference; example curl commands here are kept in lock-step with production behaviour.
Does the API support webhooks?
Lead-capture and query-completed events fire to a tenant-configured webhook URL. Signature verification uses HMAC-SHA256 with a per-tenant secret. Idempotency keys are stable for 24 hours — safe to retry on transport failure.
Are there official SDKs?
npm @@slatech/bot-sdk and pip slatech are on the platform-evolution roadmap. Until they ship, every endpoint is plain JSON over HTTPS — any HTTP client suffices. Internal helpers in SLAtech.Core/Query (record QueryCommand / QueryResult / QuerySource) document the canonical request/response shapes.

Need a Pro-tier API token?

Server-to-server bearer tokens ship with Pro and above. Email the founder.

Buyer evaluation tools

Four self-serve tools for evaluating SLAtech (or any AI chatbot vendor) without a sales call:

Eval scoreboard 200-question per-vertical methodology TCO calculator Annual savings + payback period Vendor compare-tool Filter 16 vendors by 6 criteria Vendor checklist 30 procurement due-diligence questions