Update
What's new in FinOps toolkit 14
All posts
FinOps toolkit 14 is the biggest release we’ve covered here, built around three themes: connecting AI to your FinOps data, adding optimization recommendations to FinOps hubs, and previewing a new dataset for commitment discount planning.
A Copilot Studio agent for your hub data
FinOps hubs already offers Power BI reports, real-time dashboards, and a GitHub Copilot plugin. Version 14 adds a purpose-built Copilot Studio agent template that surfaces inside Microsoft 365 Copilot as a FinOps Hub Agent. It’s aimed at the audience who doesn’t want to open a dashboard at all: project and product managers, business leaders, and finance teams who need plain-language answers to “why did this go up” or “what changed since last quarter,” not another report to learn.
The template ships with agent instructions, a schema reference so it generates valid queries against the v1_2 schema, and a catalog of common FinOps questions: enough to stand up a working agent in an afternoon. Because it queries Data Explorer using the credentials of whoever’s asking, it’s governed by your existing role-based access, not a separate grant. Someone only sees what they could already see in the underlying reports.
FinOps hubs can now tell you what to do, not just what happened
Until now, FinOps hubs told you what you spent. Version 14 adds a recommendations pipeline: it ingests Azure Advisor cost recommendations and runs a built-in catalog of Azure Resource Graph queries that surface common waste (stopped VMs, unattached disks, idle load balancers, orphaned NAT gateways, and more), landing everything in one dataset alongside Cost Management’s reservation recommendations.
The more interesting part for teams with their own conventions: every query is just a JSON file with a Resource Graph query and some metadata. Drop in your own, naming-convention checks, idle resource types specific to your estate, whatever applies, and they show up in the same dataset, styled the same way as the built-ins. Disabling a default recommendation is just deleting its file.
A simpler deployment experience
The hub deployment UI has picked up a lot of settings over the years, and it was easy to end up configuring options that didn’t apply to your setup. Version 14 consolidates the choice into one hub mode selector: None (storage only, for Power BI), Azure Data Explorer, Microsoft Fabric, or Remote Hub. Pick one, and the UI only shows the settings that actually matter for it.
Preview: which meters qualify for a commitment discount
Reservation and savings plan planning usually starts with a surprisingly hard question: which meters are even eligible, and for what term? Version 14 introduces a commitment discount eligibility dataset, a precomputed lookup sourced from the Azure Retail Prices API, that answers this per meter. It’s shipping as a preview while the team validates coverage across services, so treat it as a strong starting point rather than a final answer.
Other changes worth knowing about
- Typed metadata contracts between hub application components, plus fixes for several deployment and ingestion edge cases, including multi-currency KPI calculations and US Government cloud region support.
- Power BI and workbook fixes, including a storage report crash on long-term reservation terms and corrected Azure Hybrid Benefit reporting scope.
Where AI-assisted FinOps is headed
Putting a conversational agent in front of hub data and giving FinOps hubs a native recommendations engine are both steps toward the same goal: making cost accountability something more people in the organization can act on directly, not just something a FinOps team reports out. That’s also where Allyte spends most of its time: helping teams stand up exactly this kind of capability against their own hub, safely and to their own standards.