After a failed run
An analyst asks why the nightly revenue job did not finish. Memtro lists the recent runs, finds the failed one, quotes its error message and duration, and links to the run page in Databricks.
Databricks + Memtro
Unity Catalog tables, read-only SQL and job failures from Databricks, in plain English.
Connect a Databricks workspace with a personal access token and Memtro sees what you can see: SQL warehouses, Unity Catalog catalogs, schemas, tables and their columns, and the jobs that run there. It can run a single SELECT on a warehouse, cancelled after 30 seconds and capped at 500 rows, and list recent job runs with the error message for each failure. Nothing is written.
In Memtro
What you can do with Databricks in Memtro
Ask in plain English from chat, your editor, a scheduled job or your own agent. The model picks the right Databricks tool, reads what it needs, and answers with the source cited.
Reads
- 01SQL warehouses with state and size; a stopped one starts on the first query
- 02Unity Catalog browsed as catalogs, schemas, tables and views, with owner, format and last change
- 03A table's columns with types, comments and partition columns, and a view's definition
- 04One SELECT, WITH, SHOW, DESCRIBE or EXPLAIN at a time, cancelled after 30 seconds, at most 500 rows
- 05Jobs with their cron schedule, timezone, creator and whether paused
- 06Recent job runs with result, duration and the error message for failures, counted by outcome
In practice
Ask it like you would a colleague
Once Databricks 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.
Databricks + Memtro
Which jobs failed in the last 24 hours, and what did the error say?
What tables are in prod.sales, and what columns does prod.sales.orders have?
How many orders did we take per day last week, from prod.sales.orders?
Is the main SQL warehouse running, and who created it?
Show me the definition of the view analytics.finance.monthly_revenue
Rows without a notebook
The head of sales wants last month's orders by region. Memtro reads the columns of prod.sales.orders, writes a SELECT with a date filter and a LIMIT, runs it on the warehouse and returns the rows.
Teams
Where it fits
The teams that lean on Databricks most, a question each of them asks, and what they connect beside it.
- Data and analyticsQuestions answered from the warehouse, with the SQL shownAnalyst, Data engineer, Anybody with a question, Head of data“Revenue by segment for Q3 from the warehouse, with the SQL you used”Used alongside Postgres, Stripe, Slack, Google Drive, Snowflake and 3 more.
- EngineeringKeep decisions from evaporatingEngineer, Tech lead, On call, New joiner“Who owns the billing migration, and what did we decide about retries?”Used alongside GitHub, Linear, Jira, Confluence, Slack and 5 more.
- FinanceAsk the ledger questions in plain EnglishFinance lead, Credit control, Management accountant, Founder or CFO“Which invoices over £5k are more than 30 days overdue?”Used alongside Xero, Stripe, Postgres, HubSpot, QuickBooks Online and 10 more.
Better together
Works with
A record in Databricks 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.
- AnalyticsDatabricks with Google Analytics 4Ask your traffic questions in plain English and get the numbers back.
- AnalyticsDatabricks with Plausible AnalyticsAsk about your website traffic in plain English.
- AnalyticsDatabricks with UmamiAsk about your website traffic in plain English, without sending anybody to a dashboard.
- Finance and commerceDatabricks with XeroInvoices, bills, contacts and reports, answered in plain English with figures cited to Xero.
- Finance and commerceDatabricks with StripeCustomers, charges, invoices and subscriptions through a restricted key.
- Finance and commerceDatabricks with FellohBookings, payments, refunds, chargebacks, supplier payments and payouts from the travel payments platform.
- Work managementDatabricks with JiraJQL search, issue details and projects from Jira Cloud.
- Work managementDatabricks with Clio ManageAsk the matter: its client, its fee earner, the time on it and what is still unbilled.
- Work managementDatabricks with LinearIssues, projects and teams, synced hourly into memory.
Also in Databases and warehouses
Using something else for the same job, or both? Each connector has its own tools; connect as many as you like.
- Postgres
- MySQL and MariaDB
- SQL Server
- BigQuery
- Snowflake
- AWS Athena
- MongoDB
- ClickHouse
- Amazon Redshift
- Elasticsearch
- dbt Cloud
- Fivetran
- Neon
- Appwrite
Grist- Hasura
- Meilisearch
NocoDB- Qdrant
Typesense
Weaviate
Tinybird
Hightouch
Census- Airbyte
- Budibase
- CockroachDB Cloud
- Cube Cloud
- DataHub
kintone
Monte Carlo
OpenMetadata- Pinecone
- PlanetScale
- Quickbase
Secoda
Softr
Teable- Turso
- Upstash
Vectara- Xata
- Zilliz Cloud
Setup
How to connect
Connect withWorkspace URL and personal access token
- 1
In Databricks open your profile (top right) → Settings → Developer → Access tokens → Generate new token; give it a lifetime, since the connection stops when it expires
- 2
Copy the workspace URL from the browser, without a path, and optionally the SQL warehouse ID from SQL Warehouses → the warehouse → Connection details (the last part of the HTTP path)
- 3
Paste them into Connections → Databricks in Memtro and connect; with no warehouse ID Memtro uses the first running warehouse you can see
- 4
Ask which jobs failed overnight, or what tables the sales schema holds
Also works from
Chat in the dashboard, the OpenAI-compatible API, scheduled jobs, and any MCP client: Claude Code, OpenCode, Cursor, Claude Desktop. The Databricks tools are the same everywhere.
FAQ
Questions
Can Memtro change anything in Databricks?+
Not yet. Today every Databricks tool reads, searches or lists. When the connector gains actions, each is marked, granted by name per team and used only when a person asks for exactly that, so nothing is created, sent, updated or deleted until you say so.
Who can use a Databricks connection?+
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.
Does Memtro store my Databricks data?+
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.
How do I connect Databricks?+
Workspace URL and personal access token. In Databricks open your profile (top right) → Settings → Developer → Access tokens → Generate new token; give it a lifetime, since the connection stops when it expires.
Connect Databricks in a few minutes
Seven-day free trial, then one monthly price per organisation. Connectors, jobs and memory are never metered.