Weekly event counts
Maya Okafor, growth analyst at Acme Ltd, asked how many events per day came in last week. Memtro checked the events data source's columns, then ran one count grouped by day on Acme's own Tinybird compute and returned seven rows.
How many events per day did we get last week, straight from Tinybird?
Tinybird holds your real-time analytics data. Memtro lists your data sources with row counts, size and columns, and your pipes with whether each is published as an API endpoint. It fetches rows from a published endpoint, or runs one bounded SELECT query. When asked, and once an administrator has granted the action by name, it can run copy and sink pipes, which may overwrite a target or export to your bucket or Kafka topic, append events, permanently delete matching rows, and pause or resume copy schedules. Queries and jobs run on your own Tinybird compute and count towards your plan.
Ask in plain English from chat, your editor, a scheduled job or your own agent. The model picks the right Tinybird tool, reads what it needs, and answers with the source cited.
These tools change something in Tinybird. 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 Tinybird 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.
How many events per day did we get last week?
Which Tinybird data sources are the largest by size?
What columns does the page_views data source have?
Run the top_pages endpoint for yesterday.
Run the daily_rollup copy pipe now
Maya Okafor, growth analyst at Acme Ltd, asked how many events per day came in last week. Memtro checked the events data source's columns, then ran one count grouped by day on Acme's own Tinybird compute and returned seven rows.
Tom Reilly, data engineer at BigCo, asked Memtro to delete the events a load-test account had written. Memtro ran a dry run first and reported 48,210 matching rows, then on his go-ahead deleted them permanently on BigCo's Tinybird compute. Materialised views fed by those rows were left as they were.
The teams that lean on Tinybird most, a question each of them asks, and what they connect beside it.
A record in Tinybird 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.
In Tinybird, open your workspace → Tokens and create a static token for Memtro. Never use an ADMIN or TOKENS-scoped token.
Give it DATASOURCES:READ and PIPES:READ, for all data sources and pipes or only those Memtro may read. Only if Memtro should act, add DATASOURCES:APPEND on copy targets and event data sources, PIPES:CREATE to pause or resume copy schedules, and DATASOURCES:CREATE on data sources it may delete rows from. Copy it.
In Memtro, open Connections → Tinybird, paste it as Static token, and set API host (region) if your workspace is not on api.tinybird.co.
Ask: how many events per day did we get last week?
Chat in the dashboard, the OpenAI-compatible API, scheduled jobs, and any MCP client: Claude Code, OpenCode, Cursor, Claude Desktop. The Tinybird tools are the same everywhere.
Only when you ask, and only what it has been granted: run a copy pipe now, appending to (or, if asked, replacing) its target data source; pause or resume a copy pipe's schedule; cancel a waiting or running job; append up to 20 events to a data source; delete the rows of a data source whose column matches up to 100 given values; run a sink pipe, exporting its result to the configured bucket or topic. 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.
Static token and optional API host. In Tinybird, open your workspace → Tokens and create a static token for Memtro. Never use an ADMIN or TOKENS-scoped token.
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