Operations
The reports, checks, and syncs nobody wants to own.
Helios agents query your databases, reconcile your systems, run the nightly checks, and put the numbers where your team already looks. Describe the task once and it runs every morning without a person in the loop.
The tasks
Work your operations team can hand off
Publish the morning report
The agent runs the queries, computes the derived metrics, calls out what looks unusual, and posts a report people read instead of a dashboard nobody opens.
Each morning, post yesterday's metrics with anything unusual called out.
- clickhouse·query events for the last 24h
- postgres·join account metadata
- code·compute trends and flag anomalies
- slack·post the report to #metrics
Metrics · Mar 6
- Runs 41.2k · up 6% day over day
- Signups 128 · in line with the trend
- Error rate up 0.4pp, concentrated in one region
Run the nightly data quality checks
The agent checks the invariants you care about. When one breaks it says which rows, since when, and what ran upstream.
Nightly, run our data quality checks and alert on failures.
- postgres·run 14 invariant checks
- postgres·isolate failing rows and first occurrence
- resend·email the on-call analyst
2 checks failed
- `orders.currency` null on 41 rows since 02:10
- `accounts.plan` has 3 values outside the enum
- 12 checks passed
Reconcile billing against your own records
The agent compares what Stripe billed with what your system says it should have billed, and reports every mismatch with the invoice id.
Monthly, reconcile Stripe invoices against our usage records.
- stripe·list last month's invoices
- clickhouse·aggregate metered usage by account
- code·diff billed against recorded usage
- google-sheets·write the reconciliation tab
5 mismatches found
- 2 accounts under-billed by $1,180 total
- 3 accounts billed for a cancelled add-on
- Everything else matches
Keep the systems in sync
Product usage becomes CRM fields, spreadsheet rows, or an email. The agent handles the shape change in code and writes it where it belongs.
Nightly, sync product usage into the CRM as account health fields.
- postgres·compute 30-day usage per account
- code·map usage into health tiers
- attio·update 480 account records
480 accounts synced
- 31 moved into `at risk`
- 18 flagged as expansion candidates
Keep files and access tidy
A scheduled agent sweeps shared folders for files exposed beyond your domain and reports each one with its sharing settings.
Weekly, flag any file in `Finance` shared outside the company.
- google-drive·list files and permissions in `Finance`
- code·filter external and link-shared files
- slack·DM the finance lead
3 files shared externally
- 1 open to anyone with the link
- 2 shared with a former contractor
- Suggested revocations listed
Reach
The tools operations agents reach for
Each agent gets the complete API of every service you connect, with credentials attached server-side.
Controls
Everyone delegates. IT stays in control.
Operations agents touch your most sensitive systems, so the controls sit where they belong. Database credentials are envelope-encrypted and can reach private networks over an SSH tunnel; read-only roles make a reporting agent incapable of writing; org roles and resource visibility decide who can run or edit each task.
- SAML and OIDC SSO with SCIM provisioning
- Org roles and resource-level sharing
- Envelope-encrypted secrets, handled server-side
- SOC 2 Type II, with an isolated sandbox per agent
Use cases
Operations tasks in detail
Database reporting
An agent introspects your schema, writes the SQL, runs it, analyzes the results in its own sandbox, and posts a report your team reads — instead of a dashboard nobody opens.
Billing operations
An agent watches Stripe for failed charges and starts recovery, reconciles what you billed against what your own system recorded, and reports the revenue numbers every morning.
Weekly digests
An agent reads a week of activity across the tools your team uses, separates decisions from chatter, and posts a digest people actually read.
Templates
Start from a working agent
Every template ships with the prompt, the trigger, and the integrations it needs.
Operations
Weekly metrics report
Queries your warehouse, computes the metrics your team argues about, and posts a report with the anomalies called out.
Operations
Nightly data quality checks
Runs your invariants against production every night and tells the on-call analyst which rows broke and since when.
Operations
KPI sheet refresh
Pulls from the warehouse every morning and writes the finished tables into the sheet leadership already opens.
Operations
Drive organizer & access audit
Files loose documents by what they actually contain, and reports anything shared beyond your domain.
Sales & RevOps
Churn risk watchlist
Joins billing signals with product usage and support volume to surface the accounts quietly drifting toward the exit.
Guides
Get the first one running
10 min
Connect a database
Give an agent Postgres, MySQL, or ClickHouse. It introspects the schema, writes real SQL, and reasons over the results in its own sandbox.
8 min
Put an agent on a schedule
Turn a working prompt into a standing task. Describe the cadence the way you would say it out loud, and Helios writes the cron expression for you.
8 min
Use secrets and variables
Keep the values a task needs out of the prompt. Variables hold configuration, secrets hold credentials, and both are scoped to the level where they belong.
Give your operations team an agent.
Describe the task in plain English, attach a trigger, and let it work while you sleep.