Chat API
Send prompts and receive grounded replies with human fallback hooks.
AI & Automation APIs
AI API for chatbots, RAG search and workflow automation
Add intelligent capabilities to your product with LOOKME AI APIs — chatbot assistants, document search (RAG), workflow triggers, classification and analytics endpoints. Your software calls our models; you do not need an ML team in-house.
In Development
lookme-api · AI Chat
POST /v1/ai/chat
Authorization: Bearer <api_key>
{
"prompt": "Summarize support ticket #441",
"context": "tickets"
}
200 OK
{ "reply": "Customer asked about refund...", "tokens": 128 }
REST
Inference APIs
Tools
Function calling
RAG
Grounded answers
Beta
In development
Artificial intelligence as API capability
LOOKME exposes AI as callable API endpoints — language models for chatbots, retrieval-augmented generation for document search, classification for content moderation and workflow triggers for business process automation.
Your backend sends requests to LOOKME AI endpoints and receives structured JSON responses. We handle model hosting, scaling and inference optimisation so your engineering team focuses on product features.
AI products are in active development with early access available by enquiry. Specific model availability, latency guarantees and data handling policies are confirmed per engagement during onboarding.
Important: AI outputs can be wrong. LOOKME does not claim autonomous decision-making without review for high-risk processes. Use of third-party model providers is subject to their terms and to applicable data-protection law.
Example API endpoints
Representative REST routes issued after onboarding. Paths and fields may vary per project.
Method
Path
Description
POST
/v1/ai/chat
Send a prompt and receive a model response
POST
/v1/ai/embed
Generate embeddings for search or ranking
POST
/v1/ai/classify
Classify text into categories you define
POST
/v1/ai/workflow/run
Trigger an automation workflow step
GET
/v1/ai/usage
Token and cost usage for billing periods
AI & automation capabilities
Capabilities your engineering team integrates through REST endpoints and webhooks.
Embedding API
Vector embeddings for search, ranking and retrieval workflows.
Classification API
Route tickets, content or records into categories you define.
Workflow API
Trigger automation steps that may draft, summarise or route work.
Recommendation API
Ranking and suggestions with tunable business rules on top.
Analytics API
Narrative summaries and anomaly hints over existing metrics.
Tool calling
Let models call approved internal functions with audit logs.
Usage API
Token, cost and quota reporting per project or tenant.
Safety filters
Refusal rules, moderation flags and optional human review queues.
Webhook events
Signed HTTPS callbacks so your backend stays in sync without polling.
ai.workflow_completed
Automation run finished with output summary
ai.moderation_flagged
Content flagged by safety filters
ai.quota_warning
Usage approaching configured limit
How integration works
Typical steps from enquiry to live API access.
-
01
Use-case fit
Confirm data access, risk level and whether a model is actually required.
-
02
Grounding & safety
Define sources, refusal behaviour, logging and human review.
-
03
Build
APIs, prompts/tools, evaluation sets and admin controls.
-
04
Monitor
Quality, cost and incident handling after launch.
Integrate this API with LOOKME
Share your integration requirement, target markets and expected volumes. We will outline endpoints, webhooks and API access steps.