SuggestAPISuggestAPI

For · Typesense

Search engine

Keep Typesense, and put a discovery layer in front.

Your Typesense collection already holds the catalog and ranking. SuggestAPI routes storefront and agent requests to that collection and returns a consistent, agent-ready result shape.

Client: Storefront or agent, then Adapter: SuggestAPI, then Engine: Typesense, then Results: Ranked products

Client

Storefront or agent

Adapter

SuggestAPI

Engine

Typesense

Results

Ranked products

Typesense

Typesense stays the engine

Collections, documents, and ranking stay in Typesense. SuggestAPI does not migrate you off Typesense.

Discovery layer

SuggestAPI is the adapter

One API for autocomplete, full search, compare, and recommend. SuggestAPI rewrites the shopping job, calls Typesense, and normalizes the response for people and agents.

SuggestAPI is an independent layer in front of Typesense. It is not a Typesense product.

What stays

Leave the stack you already trust in place

  • The collection

    Documents, facets, and ranking stay in Typesense.

  • Credentials

    Typesense keys stay on the server. Clients call SuggestAPI.

  • Your cluster

    SuggestAPI calls the Typesense you already operate.

What SuggestAPI adds

Add discovery without a second catalog

  • Query understanding

    Shopping jobs and follow-ups, not only prefix strings.

  • Agent-ready results

    Structured product context for compare, recommend, and handoff.

  • Swap later if you need to

    The client stays on SuggestAPI if the backend changes.

Start with the stack you have

Keep Typesense and add discovery

Sign up, connect the catalog or search you already run, and put SuggestAPI in front.

Also for