A woman studies an AI-powered cloud data dashboard displaying analytics, database workflows, and system status.
Lucidity has launched an AI-driven FinOps tool that does more than flag cloud waste. According to Technology Record, the company launched its AI Action Engine on the opening day of Microsoft Ignite in San Francisco, which runs until 20 November 2026. The tool is built to carry cost fixes through to execution. Much of what follows is vendor-reported, so this piece attributes it and notes where claims are unverified.

A woman studies an AI-powered cloud data dashboard displaying analytics, database workflows, and system status. What Lucidity says it launched​

Technology Record reports that the platform uses AI-powered agents to help organisations cut cloud costs by identifying, validating and implementing optimisation opportunities. Lucidity's launch release says the engine is behind every Lucidity savings module. Trade coverage describes it as an automation layer added to existing savings products. So this is not a new standalone product line.

The problem it targets is familiar to anyone who has run a cloud estate. Cost-saving recommendations get identified but often go unimplemented because engineering has competing priorities. Lucidity's pitch is to remove engineers as the bottleneck.

How the workflow is supposed to work​

Lucidity's materials describe three stages: find, decide and act.

  • Find: The agents look at how resources are actually used, not only at billing data.
  • Decide: the platform weighs workload requirements and operational context to judge which changes are safe.
  • Act: Agents make the change with the user's approval, or autonomously for actions the user has preapproved. Technology Record adds that every change is monitored, verified, logged and reversible.

Lucidity's site says customers can run recommendation-only, approval-based or selected autonomous modes. It lists guardrails as approval rules, change windows, and policies set by organisation, environment, account or tag. Its examples include snapshotting before every change, skipping resources tagged to opt out, and requiring sign-off for redundancy changes.

Where Azure fits​

Microsoft Ignite is the launch venue. Nothing in the reporting shows a Microsoft partnership, Azure Marketplace listing or Microsoft endorsement. Azure is one of three supported clouds: the platform supports AWS, Azure and Google Cloud.

The Azure-specific material is concrete, though. Lucidity's current modules cover storage:

  • Disk AutoScaler predicts capacity needs, then grows or shrinks block storage.
  • Disk Tiering scores volumes against 30, 60 and 90 days of IOPS, throughput and latency.
  • Disk Cleanup finds idle, unmounted and zero-I/O disks.
  • Object Mapping, Object Tiering and Object Drift cover Azure storage accounts, including access tiers, geo-redundancy and premium-tier waste.

The site lists Compute Rightsizing as "coming soon." Don't assume VM optimisation has launched.

Lucidity's site works through an example. A 2 TiB Premium SSD data disk on a SQL Server workload had 184 GiB used and a 212 GiB peak over 30 days. The proposed change was to a 512 GiB Standard SSD, with a monthly cost drop from $259.05 to $38.40. This is a vendor illustration, not a named customer case or an independent benchmark.

The customer evidence​

Lucidity's headline figures are:

  • Customers have achieved cost reductions of up to 40%, and the company says it has identified more than $500 million in savings.
  • Presidio's chief solutions officer, Dan O'Brien, called it a "game-changer" in cloud storage management. That is a testimonial, not a test.

The most specific reference is Commvault. Per the company's release, Lucidity manages more than 3 PB across 3,500+ VMs and 11,000+ disks on Azure. Commvault has seen about 40% lower cloud storage cost and more than $2 million in Azure savings to date, with no downtime on its backup servers. Treat this as the vendor's account. The releases don't give a baseline or a measurement period.

The company's site also claims generic FinOps tools find about 20% of cloud spend in savings, against 40% for Lucidity. No methodology is given, so don't treat these as industry facts. "Up to 40%" is a ceiling, not a typical result.

Questions to ask before connecting it to production​

For Azure admins, the useful questions are about control and verification:

  1. Permissions. Lucidity says it starts read-only and uses least-privilege access through cloud APIs. Check exactly which Azure roles it requests before granting write access.
  2. Scope. Start with non-production subscriptions and apply the opt-out tag convention.
  3. Change windows. Production disks should change only inside approved windows, with snapshots first.
  4. Rollback. "Reversible" is a vendor claim. Test it on a disposable disk.
  5. Audit. Confirm the action log can be exported to your SIEM or change-management records.
  6. Data handling and compliance. Lucidity says it never reads application data and is SOC 2 Type II certified. Ask for the report rather than relying on the badge.
  7. Pricing. One third-party roundup lists a savings-share model with a free assessment. Confirm commercial terms directly.

Analysis​

The idea is sound. Autonomous changes to live production disks deserve more scrutiny than a dashboard recommendation, and the real test is how the guardrails behave on awkward workloads. Lucidity's own illustration involves SQL Server, a workload where IOPS and latency headroom matter. Shrinking and re-tiering are also not symmetric risks. Moving a disk from Premium to Standard can hurt performance if a workload spikes outside the observed window.

The independent evidence is thin. The reviewed sources contain no third-party benchmark, no failure or rollback statistics, and no audit documentation. Coverage so far is based on Lucidity's own release and site.

Bottom line​

Lucidity's AI Action Engine is an automation layer over its existing storage-focused cost modules, covering Azure, AWS and Google Cloud. Its main differentiator is automated, approval-gated execution with logging and rollback. Compute optimisation is still marked as coming soon. The savings and downtime claims are Lucidity's own. A read-only assessment on a non-production scope is the sensible way to test them.

 

References

  1. Microsoft Ignite 2026: Lucidity launches the AI Action Engine - Technology Record Technology Record 2026-11-17T20:00:00+00:00
  2. Lucidity Launches the AI Action Engine to Turn Cloud Savings Into Delivered Results manilatimes.net
  3. Lucidity launches India-built AI engine to automate enterprise cloud savings dqindia.com