Today, we are introducing the CloudFlow SQL node – a way to turn the data you already have into live FinOps and CloudOps automations, using the language your teams already know: SQL.

Instead of exporting data to a BI tool, building a dashboard, and hoping someone remembers to check it, you can now write a SQL query inside CloudFlow and wire the result directly into alerts, workflows, and policies across all of your clouds. The CloudFlow SQL node natively connects to our unified data lake, using the same normalized schema you already use in DoiT Cloud Intelligence™.

While we’re excited to see what you build using the SQL node in CloudFlow, we've helped get you started with several pre-built templates available in the CloudFlow template library (use the Source: Billing Data filter to quickly find them). 

These templates also require no configuration to start running them. Simply click one to open it, and hit publish to start generating meaningful alerts and workflows. Read on for more details about one example you can start implementing:

Example: Time-to-Tag Leakage Analysis (Tag Lag Drift)

Here's a common FinOps problem: even when teams consistently tag resources, tags often appear days after the resource starts incurring costs. The spend generated during that gap shows up as “untagged,” creating leakage that can’t be attributed to the correct owner or workload. Dashboards can display daily untagged spend, but they do not quantify the lag effect or identify services where the lag is worsening.

1. Scans billing data for each resource and computes the first timestamp where a non-system tag is present
2. Compares that timestamp to the resource’s earliest usage cost and calculates leakage per resource, per service, per day
3. Aggregates to daily untagged leakage, total untagged spend, and the percent of spend impacted by tagging delay
4. Computes trend direction (improving or degrading) over a defined window
5. Filters out services that do not support cost allocation tags to avoid noise

In CloudFlow, you can then:

• Trigger an alert only when leakage crosses a threshold or when the trend degrades week over week
• Notify owning teams of the specific services or resources driving the lag (or create a Jira task)
• Feed the data into an CloudFlow's LLM node to generate a human-readable explanation of what caused the degradation and which teams should act

This converts “some of our spend is untagged” into an operational metric with direction and ownership.

What you can build next

This example is only a starting point. With the CloudFlow SQL node, you can also design automations for:

• Budget guardrails and burn rate monitoring at the project, account, or business unit level
• Policy checks for label compliance are tied directly to your own label taxonomy
• Identification of noisy SKUs, high-cost regions, or underutilized data services
• Custom CloudOps SLOs driven by spend and usage patterns, not just uptime

If you already use DoiT Cloud Intelligence, the CloudFlow SQL node is available in your environment as a new node type inside CloudFlow. Start by taking one of your existing FinOps queries and turning it into an automated Flow that runs on a schedule, posts to Slack, opens Jira tickets, or chains into additional nodes such as LLM-based explanations.

Learn more about the SQL Node in DoiT Cloud Intelligence Help Center. We can also help you build your automation free of charge! Just ask at support.doit.com.

If you’re using Azure AI to build and run LLM-powered applications, you’ll now see those associated costs in GenAI Intelligence alongside any AI spend from other platforms you use.

GenAI Intelligence gives you a single, comprehensive view of AI costs and token usage across your AI stack, including other supported providers like Amazon Bedrock, OpenAI, and Anthropic Claude.

Azure AI usage will also populate GenAI labels, so you can easily build your own reports and allocations on top of this AI spend data. GenAI labels turn provider-specific data into consistent dimensions (like Model, Feature, Media Format) across your AI stack, making it much easier to break down GenAI costs and usage without digging through SKUs and services for each platform.

To get started, explore GenAI Intelligence and use GenAI system labels to explore and allocate your GenAI spend.

GenAI Intelligence is available on all DoiT Cloud Intelligence™ tiers.

We’re excited to roll out a major upgrade to how DoiT helps you stay in control of your Google Cloud spend: real-time anomaly detection for Google BigQuery on-demand. For the first time, you’ll receive alerts about unexpected BigQuery cost spikes in under an hour.

What’s new

Until now, anomaly insights for BigQuery relied on next-day billing-file ingestion. That meant if a bad query ran at 2 PM today, you wouldn’t know until tomorrow.

With our new real-time detection pipeline for BigQuery on-demand, we continuously ingest and analyze live BigQuery usage metadata, flag unusual usage patterns, and send you Slack or email alerts in less than an hour – not the next day.

Why it matters

As DoiT’s BigQuery expert Sayle Matthews will tell you, the risk of unchecked queries racking up significant costs in a short period of time is very high:

“One of the largest issues is seeing how much their bill is at any given moment and being able to alert them when a ‘runaway query’ hits. We have seen some examples where customers have single queries that cost $2,000 USD and run in less than a minute, and of course, these were run multiple times in quick succession. These mistakes lead to massive bills that aren't caught for days or weeks later.”

Real-time detection for BigQuery on-demand means that you can:

• Catch runaway queries in minutes: Prevent accidental or inefficient queries from racking up costs before anyone notices.
• Protect against operational mistakes: Get alerted when abnormal query activity starts impacting spend.
• Strengthen your security posture: Real-time cost changes can signal unauthorized data access or compromised systems.

What you need to do

This feature is available for all DoiT customers with a paid Enhanced, Premium, or Enterprise subscription and a connected Google Cloud account. To enable it, take the following steps:

1. Locate the service account of interest on the Google Cloud access & features page.
2. Select the kebab menu (⋮) next to the project connection, and then select Edit.
3. Select the Real-time Anomalies – BigQuery checkbox to add the feature.
4. Select Generate gcloud commands.
5. Follow the instructions displayed in the side panel to update your custom role.
6. Select Done to enable the feature

Enable BQ real-time anomaly detection

Next steps

Up next, we’ll be releasing the same support for BigQuery reservations to bring the same real-time intelligence across your full BigQuery footprint.

For more information about enabling real-time anomaly detection, consult our Help documentation or raise a support ticket.

For customers sending (or planning to send) Snowflake cost & usage data to DoiT Snowflake Intelligence, key-pair authentication is now supported and required for new connections, and existing password-based LEGACY_SERVICE users must migrate by June 2026.

Snowflake is deprecating password-only LEGACY_SERVICE users, with full removal planned for June 2026. In response, we’ve added support for key-pair authentication. This removes passwords from the flow and gives you stronger security and a cleaner audit trail for your Snowflake connection.

What this means:

• Starting today, key-pair authentication is the only way you’ll be able to set up the Snowflake Intelligence
• If you’ve already set up the integration with a LEGACY_SERVICE user, migrate to key-pair authentication before June 2026 to avoid interruption

To get started, view our documentation on setting up and updating your Snowflake connection.

You can now use DoiT's Agentic FinOps AI, directly from any Insight detail page to get guided analysis and ready-to-use actions without leaving the view.

• Estimate the impact of implementing the Insight, including cost savings and business or technical implications based on the context.
• Break down the remediation tasks with a step-by-step plan tied to the affected resources.
• Surface downstream dependencies & hidden costs, callouts for related risks or follow-up work to expect.
• Estimate effort to execute, with time/complexity guidance that’s persisted in the UI for later reference.
• Generate a Terraform configuration, then one-click copy option for the snippet to streamline applying the fix.

Why it matters:

• Cut MTTR by getting impact, steps, and effort in seconds.
• Reduce handoffs and tabs. Everything you need lives beside the evidence.
• Improve consistency. AI gives repeatable plans and Terraform you can reuse.

No configuration changes are needed. To get started, select the AI bubble in the bottom right corner of any Insight detail page.

MongoDB provides teams with powerful elasticity, but that flexibility often comes with fragmented visibility into what drives costs. MongoDB Intelligence brings structure to that chaos.

This new module ingests Atlas billing and usage data across organizations, projects, and SKUs, turning raw cost exports into an actionable view of your Atlas estate. You can instantly see how spend breaks down across clusters, backups, and storage, identify which projects are trending up or down, and trace deltas to specific Atlas SKUs like ATLAS_AWS_INSTANCE_* or ATLAS_BACKUP_*.

Analytics for organization-level costs, project distribution, and SKU analytics reveal patterns previously hidden in CSVs, helping FinOps and engineering teams align on the same data when evaluating scaling, tiering, or retention decisions.

On the right, you’ll find an Agentic FinOps AI assistant. Ava isn’t just a chatbot; it’s a reasoning layer that interprets the same data visible on the dashboard. Ask questions like “Which projects or SKUs changed the most this month?” or “Where am I overspending on backups?”, and Ava will analyze cost trends, isolate anomalies, and suggest next best actions.

Setting up MongoDB Intelligence takes only a few minutes — connect your Atlas organization, and DoiT Cloud Intelligence™ will automatically ingest cost and usage data. Follow the step-by-step guide in our Help Center article to enable the integration securely and start visualizing your MongoDB Atlas spend with zero manual exports.

MongoDB Intelligence extends the DoiT Cloud Intelligence™ platform’s FinOps coverage to Atlas, providing your teams with a precise, explainable, and actionable understanding of MongoDB spend. MongoDB Intelligence is available on all DoiT Cloud Intelligence™ tiers

We’re excited to launch a new enhancement to Cost Anomaly Detection that gives you more control over the alerts you receive.

Until now, anomaly detection has automatically surfaced unexpected cost spikes. Today, you can fine-tune the experience with a new Sensitivity setting: a simple drop-down that lets you choose how aggressively we surface anomalies for each service in your cloud environment, including a change log that allows you to see and review all changes.

These settings will also be visible when you open a specific anomaly report:

Why this matters

Different services have different tolerance levels for noise. With sensitivity controls, you can now tailor alerting to match your operational needs – whether you want to catch every small fluctuation or focus only on the most material cost movements.

What you can do

• Adjust sensitivity levels to determine how strict anomaly thresholds should be
• Reduce alert noise in high-variance environments
• Prioritize only high-impact events when cost stability is strong
• Dial up sensitivity when monitoring new workloads or emerging spend patterns

Getting started

Head to your anomaly detection settings and select the Sensitivity option that best fits your use case. No configuration or data prep required, just choose from the new drop-down and your anomaly detection model adjusts automatically.

To learn more about the sensitivity controls, check out our Help documentation.

You can now track detailed costs and token usage for Google Vertex AI and Databricks GenAI workloads in GenAI Intelligence. GenAI Intelligence gives you a single, comprehensive view of AI costs and token usage across your stack, including other supported providers like Amazon Bedrock, OpenAI, and Anthropic Claude.

For Databricks, coverage includes:

• Databricks-hosted foundation models
• Custom models on Mosaic AI Model Serving
• Model serving–related network egress
• SQL AI Functions usage

Additionally, cost and token data for these AI providers will populate in GenAI system labels. These system labels standardize your GenAI data into consistent categories (ex. Model, Usage Type, Media Format) across all supported GenAI providers, making it easier for you to drill down into your GenAI costs and usage across providers without going through service-specific SKUs and resources.

To get started, explore GenAI Intelligence and use GenAI system labels to report on and allocate your GenAI spend.

Note: In order to view Databricks GenAI workloads in GenAI Lens, you’ll need to first connect your Databricks account to DoiT Cloud Intelligence.

Tracking and monitoring commitments can be hard, especially when you’re juggling multiple contracts and need quick, consistent updates. 

We’ve added two ways to keep commitment progress front and center without needing to log in or dig through reports.

What’s new:

• Subscribe to commitment updates. Receive a regular summary of your commitment’s progress via email or Slack, including spend to date, remaining value, and forecast. Set your preferred schedule, time zone, and frequency. The digest keeps you continuously informed, automatically.
• Add commitments to dashboards. Gain visibility where you need it by adding any Commitment Manager widget directly to your dashboards. Track spend and performance alongside your other key metrics, and jump into the full commitment view with one click.

To get started, go to Monitor → Commitment Manager, or Dashboards → Add Preset Widget.