Odd search results
The product manager at Northwind asks why a search for desk lamp shows chairs first. Memtro runs the same search, shows the 214 matches, and the schema shows the sorting field is popularity.
Ask what a Typesense search for a phrase returns, and how each collection is shaped.
Connect Typesense with one node's host and an API key and Memtro reads collections with their document counts and schemas, aliases, searches a collection and reads one document. Typesense is self-hostable, or runs on Typesense Cloud. Limit the key to the collections Memtro should see. When you ask, and once an administrator has granted them by name and the key allows, it can create, update, import or delete documents, create, alter or drop collections, repoint aliases, and manage synonyms, curations and stopwords. Deleting documents, a collection or a set is permanent, and overwritten values cannot be recovered.
Ask in plain English from chat, your editor, a scheduled job or your own agent. The model picks the right Typesense tool, reads what it needs, and answers with the source cited.
These tools change something in Typesense. The model is told to use them only when you ask for exactly that, and an administrator grants each one by name to the teams that may use it; a personal connection acts only after its owner allows it.
Once Typesense is connected, these are the kinds of things people ask. Plain questions, no syntax. If the answer lives partly in another system you have connected, Memtro looks there too.
What does a search for wireless charger return in products?
Which fields in the products collection are facets?
How many documents are in the help-articles collection?
Point the products alias at products_v2
Delete every document in products where discontinued is true
The product manager at Northwind asks why a search for desk lamp shows chairs first. Memtro runs the same search, shows the 214 matches, and the schema shows the sorting field is popularity.
Dan Fletcher has reindexed the Acme catalogue into products_v2 and asks Memtro to switch the site over. Memtro checks both collections' counts, then repoints the products alias, and every app searching through it reads the new collection at once.
The teams that lean on Typesense most, a question each of them asks, and what they connect beside it.
A record in Typesense is rarely the whole story. Connect the systems that hold the rest and Memtro checks them together, then remembers what it learned so the next question starts further along.
Using something else for the same job, or both? Each connector has its own tools; connect as many as you like.
Make a key for Memtro (Typesense Cloud → Cluster → Generate API Keys, or POST /keys with the admin key) with collections:list, collections:get, documents:search and documents:get, limited to the right collections. Add documents:create, documents:upsert, documents:update, documents:import and documents:delete, or collections, aliases, synonym and curation actions, only where Memtro may change them
Type one node's host name only, such as xyz.a1.typesense.net, no https:// or path: a public DNS name with a publicly trusted certificate. Plain http on 8108 is refused. Set the port if https is not on 443
Paste the API key and the Host into Connections → Typesense in Memtro. A search-only key fails the connection test
Ask what a search for a phrase returns
Chat in the dashboard, the OpenAI-compatible API, scheduled jobs, and any MCP client: Claude Code, OpenCode, Cursor, Claude Desktop. The Typesense tools are the same everywhere.
Only when you ask, and only what it has been granted: create a document in a collection; upsert a document, replacing whole any document with the same id; update fields of one document; delete one document; delete the documents in a collection that match a filter; import documents in bulk (create, update, upsert or emplace); create a collection; delete a collection with all its documents; change a collection's schema: add or drop fields, link synonym or curation sets; update the same fields on up to 100 listed documents; create or repoint an alias to a collection; delete an alias; create or replace a synonym set, or one synonym in it; delete a synonym set or one synonym; create or replace a curation set, or one curation (pin or hide results for a query); delete a curation set or one curation; create or replace a stopwords set; clear the search result cache. Every other tool reads, searches or lists. The Connections page labels the tools that act, an administrator grants each one by name to the teams that may use it, and a personal connection acts only after its owner allows it.
A personal connection serves only your own requests. A shared connection serves the organisation: everyone, or only the teams an admin chooses. Admins can always add shared connections; other members need the permission.
Credentials are encrypted at rest. Tool results are cached briefly so agent loops do not repeat identical calls, and the facts the model extracts are saved to your organisation's memory with provenance (system, record and link) so they can be checked and superseded.
Host and API key. Make a key for Memtro (Typesense Cloud → Cluster → Generate API Keys, or POST /keys with the admin key) with collections:list, collections:get, documents:search and documents:get, limited to the right collections. Add documents:create, documents:upsert, documents:update, documents:import and documents:delete, or collections, aliases, synonym and curation actions, only where Memtro may change them.
Seven-day free trial, then one monthly price per organisation. Connectors, jobs and memory are never metered.
Sign in, add a provider key, create an API key and point your tools at Memtro. It takes about five minutes.
7-day free trial · card on file, nothing charged until it ends · cancel any time