AI in Finout (Alpha)

Overview

Finout AI for FinOps sets a new standard for FinOps solutions: it uses AI to build and maintain cost tags, allocation rules, and spend breakdowns and slashes manual effort to a minimum - so every team, from FinOps and finance to engineering, gets fast, intuitive answers on cloud spend. AI-Built Allocation, Human Review:

• AI suggests tags, rules, and groupings; users review, edit, and approve.

• Clear explanations (what pattern was found, what resources are affected)

From 80% to 100% Cost attribution:

• Attribute all your unallocated resources across clouds, K8s, and SaaS.

• Better allocation → better showback/chargeback, unit economics, and forecasting.

FinOps for Everyone:

• Natural language questions instead of query syntax, so anyone can get to answers quickly.

• Finance, product, and engineering can self-serve cost insights.

• Supports FinOps Foundation goals around shared accountability.

Self-Improving FinOps:

• Suggest new or improved options as your environment changes.

• Learns from your approvals/edits so it gets closer to how you think about teams, products, and environments.

Note: AI never uses customer data for model training, and all processing stays securely isolated within your workspace.

Prerequisites

AI in Finout requires two prerequisites:

1. Admin: An Admin must opt in to AI Services for the workspace.

   • Performed in Settings → Compliance & Privacy → AI Services. See AI Opt in for more information.

   • Opt-in to enable AI capabilities across the entire account.

   • Recorded in the audit log. AI Opt-In - Finout Docs

2. User: An Admin must enable AI permissions

   • Certain AI-powered features require feature-specific activation for your users.

   • Admins maintain full control over enabling and disabling these capabilities.

Finout Features with AI

AI Filter

Enables you to ask cloud cost questions in plain language and get clear breakdowns and charts - no query language or dashboard building required.

• Translates natural language questions like “What did we spend on production AI workloads last month by team?” into real queries on your MegaBill/Data Explorer.

• Returns tables and charts that users can refine with follow-ups like “only Kubernetes” or “compare to last quarter.”

• Works for FinOps, finance, product, and engineering—no training, no syntax.

Virtual Tags with AI

AI Virtual Tags

AI Virtual Tags automate cloud, Kubernetes, and SaaS cost allocation - replacing manual tagging with AI-generated rules you can approve in a click. This helps you:

• Eliminate tagging backlogs – Replace months of tagging cleanup with AI-suggested rules you can approve in minutes.

• Get to 100% allocation faster – Achieve full coverage across cloud, Kubernetes, and SaaS without waiting on engineering.

• Spend time on optimization – Focus on anomalies, unit economics, and commitments instead of wrangling tags.

AI Virtual Tag Rule Builder

Finout AI Rule Builder turns plain-language intent into precise allocation rules, so maintaining your cost model feels like editing text- not writing filters and SQL. This can be used for any new virtual tag or existing virtual tag.

• Converts a simple request like “Group all ‘-dev’ resources under Dev Environment” into a complete rule using names, tags, and labels.

• Suggests rules to fix unknown or misallocated spend when new services, accounts, or clusters appear.

• Keeps rules fully editable, so FinOps can adjust logic without starting from scratch.

FAQs

Who can enable AI features?

Only Admins can enable or disable AI Services from Settings → Compliance & Privacy.

Does Finout use my data to train AI models?

No. Finout’s AI features never use your data for model training, and all information remains isolated within your workspace.

What happens when I disable AI Services?

Finout immediately disables all AI-powered features for the workspace, and the action is logged.

What happens to my filters and settings while the AI is processing my request?

All existing filters, groupBy, timeframe, and interval fields remain visible while the AI processes your prompt. This ensures you maintain full context and can continue adjusting, editing, or canceling the configuration at any time without waiting for the AI to finish.

What are the prompt input requirements?

AI prompts must contain at least three words and be no longer than 100 characters. This enables Finout to have sufficient context to understand your intent and reliably generate a valid configuration for filters, grouping, and timeframes.