Analyze your brand, automate it in 2 minutes
Azure Cosmos DB

Azure Cosmos DB

#342 most-used

Query, write, and manage your NoSQL data store

StorageDatabaseAnalyticsDeveloperAutomationCloud & Infrastructure

Azure Cosmos DB is Microsoft's globally distributed, multi-model NoSQL database designed for applications that demand low latency and high availability at any scale. It stores JSON documents in schema-free containers, supports a SQL-like query language, and is partitioned for planet-scale throughput. Connect it to Actionist and your agents can create and manage containers, write and update structured JSON items, execute parameterised SQL queries to pull exactly the records you need, and delete documents with full audit trails — all without anyone touching the Azure portal.

Try Azure Cosmos DB with ActionistVisit Azure Cosmos DB
Average time saved
14 hours
per person · per month
≈ 2 workdays back

Eliminates manual work. Agents eliminate manual database queries, exports, record creation, and audit sweeps — work that previously required engineers or analysts to touch the Azure portal or write ad hoc scripts.

Schedule

What your Azure Cosmos DB agent runs on autopilot

A week of scheduled jobs your Actionist agent will execute on your behalf.

28Scheduled jobs
7Agents at work
24/7Always on
Agents
Wed–Fri
Wed
Thu
Fri
7a
8a
9a
10a
11a
12p
1p
2p
3p
4p
5p
6p
Multi-app workflows

Azure Cosmos DB × every other app you use

End-to-end automations that span multiple apps — each one a real business outcome.

6Workflows
8Apps spanned
~25 hrsSaved / week
6Personas served
For sales
Featured4 apps

Deal stage change triggers Cosmos DB record update

When a deal moves to Proposal Sent in HubSpot, the agent reads the prospect's profile from Cosmos DB, updates the deal document with the new stage and date, posts a proposal summary to #deals, and schedules a 48-hour follow-up reminder — keeping the data store and the team in sync without manual entry.

~5 hrs

Time saved for your team — every week, on autopilot

The flow
Trigger·When a deal advances to Proposal Sent stage in HubSpot
Trigger
Step 1
HubSpot
Detect deal stage change to Proposal Sent
read
Step 2
Azure Cosmos DB
Get the prospect's profile document from the leads container
write
Step 3
Azure Cosmos DB
Update deal record with proposal_sent_date and stage field
write
Step 4
Slack
Post proposal summary to the #deals Slack channel
write
Step 5
Google Calendar
Schedule a follow-up reminder in 48 hours
Result
Update deal record with proposal_sent_date and stage fieldPost proposal summary to the #deals Slack channelSchedule a follow-up reminder in 48 hours
The win
Saved per run
15 min
Runs / week
~20×
Every deal stage change is captured in the operational data store automatically
Driven bySales Agent
ROI

Savings

What your team gets back — two angles: what you stop doing manually, and what that's worth.

Without Actionist

What you do manually today

With Actionist

What your agent runs for you

  • Sales
    25 min / week
    Manual deal record maintenance

    Reps copy deal data from HubSpot into the internal database manually, running queries through the Azure portal each time they need a report.

    Sales Agent
    0 min
    Agent writes and queries deal records automatically

    When a deal advances in HubSpot, the agent updates the Cosmos DB record, queries for related prospects, and posts a summary to Slack — all without the rep touching the database.

  • Marketing
    90 min / week
    Manual audience data export

    Marketing requests data exports from the engineering team, waits for a data engineer to run the query, and manually formats the output into a spreadsheet for campaign planning.

    Marketing Agent
    0 min
    Agent queries audience segments on demand

    The marketing agent runs segmentation queries against Cosmos DB on a schedule, exports the results to Google Sheets, and posts a briefing — no analyst or data engineer needed.

  • Customer Support
    60 min / week
    Manual customer history lookup

    Support agents search for the customer in the database, open their profile in a separate tab, and manually copy relevant context into the ticket before responding.

    Customer Support Agent
    0 min
    Agent enriches every ticket with customer history at creation

    When a ticket is created, the agent reads the customer profile and prior cases from Cosmos DB and adds them as internal notes — agents start every case with full context.

  • Human Resources
    20 min / week
    Manual onboarding record creation

    HR manually creates database records for each new hire, copying details from the HRIS into the internal system and updating checklist status throughout the onboarding period.

    Human Resources Agent
    0 min
    Agent creates and maintains onboarding records automatically

    When a new hire is confirmed, the agent creates a structured onboarding document in Cosmos DB with all checklist items, queryable by IT, HR, and the hiring manager from day one.

  • Finance
    90 min / week
    Manual transaction export and reconciliation

    Finance manually exports transaction data from the database, copies it into a spreadsheet, and runs SUMIF formulas to calculate product-line totals for the weekly review.

    Finance Agent
    0 min
    Agent runs the weekly reconciliation from Cosmos DB automatically

    The finance agent queries the transactions container, aggregates totals by product line, and writes the summary to Google Sheets and the reporting container — before the Friday board review.

  • Operations
    30 min / week
    Manual container inventory and audit

    Operations engineers log into the Azure portal weekly, browse each database's container list, and manually compare it against a spreadsheet to catch schema drift or shadow containers.

    Operations Agent
    0 min
    Agent inventories containers and flags anomalies weekly

    The operations agent lists all containers, cross-references them against the schema registry, and flags any undocumented ones to #ops-alerts — without anyone opening the Azure portal.

  • Legal
    45 min / week
    Manual contract deadline tracking

    Legal operations staff manually query the contracts database, cross-reference deadlines against a calendar spreadsheet, and email reminders to the responsible attorneys for each upcoming review.

    Legal Agent
    0 min
    Agent monitors contract and compliance deadlines automatically

    The legal agent runs weekly queries for upcoming review dates and overdue compliance items, posts summaries to #legal, and updates records with deadline flags — no manual calendar checks.

+ 100s of other Azure Cosmos DB automations
Average time saved
36 hrs / person / month
Calculator

Calculate what your team saves

Team size
8 people
Hourly rate
$30 / hr
Hours saved / week
28
Hours saved / year
1,400
Annual ROI
$42,000

Based on Azure Cosmos DB's typical team usage — the visible tasks plus a few other automations the agent runs: ~3.5 hrs / person / week of admin work automated.

Connect

How to plug Azure Cosmos DB into Actionist

Pick the connection method that suits your environment.

Connect using your Azure Cosmos DB account name, a primary or secondary account key, and the target database name. All three values are available in the Azure portal under your Cosmos DB account.

1
Find your keys in the Azure portal

In the Azure portal, navigate to your Cosmos DB account and select Settings > Keys. You will see your account name and two primary/secondary keys.

2
Copy the required credentials

Copy your account name, one of the account keys (primary or secondary), and the name of the database you want to connect to.

3
Paste into Actionist and test

In Actionist, find Azure Cosmos DB in the Apps tab, click Connect, paste the account name, key, and database name, and click Test connection. Actionist runs a read-only query to confirm the handshake.

Credentials you'll need
Account Name*
The name of your Azure Cosmos DB account (found in the Azure portal overview).
Account Key*
A primary or secondary key for your account. Found under Settings > Keys in your Cosmos DB account in the Azure portal.
Database*
The name of the Cosmos DB database to connect to.
Actions

12 actions your agent can call

Read and write operations available to your Actionist agent.

FAQs

Questions about Azure Cosmos DB + Actionist

How does Actionist connect to Azure Cosmos DB?
Go to the Apps tab in Actionist, find Azure Cosmos DB, and click Connect. You will need three pieces of information from your Azure portal: your Cosmos DB account name, one of your account keys (found under Settings > Keys in your Cosmos DB account), and the name of the database you want to connect to. The agent runs a read-only test query against the database to confirm the connection before any actions run.
Which Cosmos DB key should I use, and how do permissions work?
Your agent can read and write items and containers using your account key. For production workloads, Microsoft recommends using the read-only key for agents that only need to query data, and the primary key only for agents that create, update, or delete records. You can rotate keys in the Azure portal at any time; just update the credential in Actionist after rotating.
Does Actionist support Cosmos DB's SQL query API?
Yes. Actionist supports Cosmos DB's SQL API (also called Core API), which is the most widely used. The agent can execute parameterised SQL queries via the Execute Query action to retrieve documents matching complex conditions across any container. If your account uses other APIs (MongoDB, Cassandra, Gremlin, Table), those endpoints are not natively supported through this integration, but you can reach them via Actionist's HTTP Request capability.
What are the most common automation patterns with Azure Cosmos DB?
The most common patterns are: (1) writing structured data from other apps into a Cosmos DB container as the agent processes it; (2) querying a container on a schedule and posting results to Slack or Google Sheets for reporting; (3) creating containers automatically when new projects or tenants are provisioned; (4) updating item fields when a record changes in a CRM, ERP, or support system; and (5) syncing Cosmos DB data into analytics pipelines by reading items in bulk and forwarding them downstream.
Are there throughput limits I should know about when running agents against Cosmos DB?
Yes, but with a planning note. Cosmos DB's throughput is measured in Request Units (RUs). An agent running bulk operations such as Get Many Items or Execute Query on large containers can consume significant RUs quickly. For high-volume operations, use the Execute Query action with selective projections and filters rather than fetching all items and filtering in Actionist. Schedule heavy read jobs during off-peak hours to avoid contending with your application's own throughput budget.
Can my agent write arbitrary JSON documents to Cosmos DB?
Yes. The Create Item action accepts any valid JSON document. The agent can receive a structured object from another step, for example a CRM record, a form submission, or a processed webhook payload, and write it directly to Cosmos DB. You do not need to pre-define a schema; Cosmos DB is schema-free and will accept the document as-is. The only required field is the partition key property your container was configured with.
Can Actionist react to new records written to Cosmos DB in near real time?
Actionist does not receive push events from Cosmos DB natively. Instead, the agent polls Cosmos DB within about a minute using scheduled tasks, for example an Execute Query action that checks for new items created after a stored timestamp. For near-continuous monitoring, you can run the poll on a short schedule. If you need real Change Feed integration, wire Cosmos DB's Change Feed through an Azure Function or Event Hub and use that as the trigger source for the agent.
How does deletion work, and can I set up approval before items are deleted?
The Delete Item action requires the item's ID and its partition key value. The agent will hard-delete the document from the container. Cosmos DB does not support a soft-delete natively, so if you need an audit trail before deleting, have the agent call Get Item first, log the document to a separate archive container or Google Sheets, and then call Delete Item. For high-stakes deletions, configure the agent to use an Approval Mode so a human confirms each deletion before it runs.
Get started

Connect your apps in minutes.

Start with a free instant demo, or talk to our team about a deployment designed for your business.

Get a free instant demoBook a call