€/AIObservability · Cloud & AI cost
Every cloud and AI bill. One view.
Blis Cloud & AI Cost Manager brings the costs of every cloud and AI provider you use into one open model in Elastic, next to how much you actually use.
Built on the open FinOps standard FOCUS
Where did last month's cloud and AI budget go, and which AI provider gives you the most for it?
Why now
AI spend is now part of the bill
- 98%
of FinOps practitioners now manage AI spend.
Source: FinOps Foundation, State of FinOps 2026 (opens in a new tab)
- 72%
of organisations using AI services had an unexpected AI cost spike or bill in the past year.
Source: Harness, 2026 State of AI in FinOps (opens in a new tab)
- 52%
say no single person clearly owns AI cost.
Source: Harness, 2026 State of AI in FinOps (opens in a new tab)
Is this you?
Three things we hear about cloud and AI cost
- 01
“Four providers, four invoice formats, no total.”
Cloud and AI costs arrive in different shapes from different places. Nobody sees the whole bill.
- 02
“The AI bill went up, and nobody knows why or what it is being spent on.”
Without usage next to cost, a spike is a mystery until the next invoice.
- 03
“Which AI provider is actually worth it?”
Price per token says little. What matters is what you get for the money.
What you get
Total, split, who and why
01Total
All your cloud and AI spend in one place, in one open data model.
02Split
Cloud versus AI, per provider, per service, per model.
03Who
Showback per team, workload or cost centre, from your tags and mappings.
04Why
Cost next to usage: tokens per model, seats and active users, accepted suggestions. Spikes are flagged, and an AI agent answers questions such as “Why did our AI spend rise last week?”
Alertsto Microsoft Teams, e-mail or your service desk
Monthly cost reportin the Run service
How it works
One open model for every provider
Your bills and usage
- Azure
- AWS
- Google Cloud
- Oracle Cloud Infrastructure
- Anthropic
- OpenAI
- GitHub Copilot
- Azure AI Foundry
- OpenTelemetry, as the base for others
FOCUS
Native FOCUS exports from the clouds; converters for AI and SaaS sources.
Elastic, yours
In your own Elastic, a customer-owned EU deployment, or self-managed.
- Total
- Split
- Who
- Why
- Alerts
- AI provider comparison
- AI agent
FOCUS, the FinOps Open Cost and Usage Specification, is an open standard from the FinOps Foundation. The major clouds export it natively; for AI and SaaS providers we convert their usage and cost data to the same model, so one set of dashboards works for every provider.
IllustrativeSupported sources
The clouds and AI services we cover
Native FOCUS export
Clouds
- Microsoft Azure
- Amazon Web Services
- Google Cloud
- Oracle Cloud Infrastructure
Via Blis converters
AI and SaaS
- Anthropic
- OpenAI
- GitHub Copilot
- Azure AI Foundry
Ask us
On request
Other FOCUS-compatible providers, and new sources added with our conversion tooling, with OpenTelemetry as the base
Cost data from cloud providers arrives through their periodic billing exports. AI services bought through a cloud, for example Azure OpenAI, Amazon Bedrock or Google Vertex AI, appear in that cloud's export.
AI provider comparison
Compare AI providers on what you get, not just what you pay.
We run one fixed set of questions, your own use case, against the providers and models you consider, and score every answer on correctness, latency and cost per answer. You see which provider gives you the most for your money, for your work. Available as part of the Assess step, and repeatable in the Run service.
One fixed set, from your own use case
The ones you consider
ABC- CorrectnessIs the answer right?
- LatencyHow long does it take?
- Cost per answerWhat does one answer cost?
The result is one view per provider and model, on your work: which one answers best, how fast, and at what price per answer.
IllustrativeHow to start
Scan, assess, implement, run
- 0
Scan
free · 1 hourWhich clouds and AI providers you use, who owns the cost, whether FOCUS exports are on. A short findings note.
- 1
Assess
fixed price · about two weeksA baseline on your own bills: total and split, top cost drivers, allocation gaps, cost per token and per seat, quick wins. Optional: an AI provider comparison on your own questions.
- 2
Implement
fixed price per sizeExports and converters, dashboards, alerts, the AI agent and the comparison view, set up and accepted when totals match your invoices.
- 3
Run
monthlyWe keep sources and mappings current, tune alerts, explain changes and send a monthly cost report.
Deployment and data
Your Elastic or ours, never someone else's SaaS
Your Elastic or ours
In the Elastic you already run, in a customer-owned Elastic Cloud deployment in the EU, or self-managed in your own environment.
What we process
Billing and cost data, and usage metadata: tokens, models, seats, activity. Not your prompts or AI responses.
People
User identities in seat and usage data can be pseudonymised if required, and mapped to teams.
Agreements
A data processing agreement is part of Assess and Run.
Why Blis Digital
One open model, in your own Elastic
- 01
One open model
FOCUS for cloud and AI. No proprietary schema, no lock-in.
- 02
In your own Elastic
EU or in-house; no extra SaaS.
- 03
Cost next to usage
Tokens, seats, suggestions, and your observability data.
- 04
AI value, measured
Provider comparison on correctness, latency and cost per answer.
- 05
Elastic Select Partner
Blis Digital is an Elastic Select Partner.
FAQ
Questions we get asked
A pack on Elastic that brings the costs of your clouds and AI providers into one open model (FOCUS), shows cost next to usage, per team, with alerts, an AI agent and an AI provider comparison.
The FinOps Open Cost and Usage Specification, an open standard from the FinOps Foundation that gives every provider's bill the same columns. Azure, AWS, Google Cloud and Oracle Cloud Infrastructure export it natively.
Azure, AWS, Google Cloud and Oracle Cloud Infrastructure through their FOCUS exports, and Anthropic, OpenAI, GitHub Copilot and Azure AI Foundry through our converters. Other FOCUS-compatible providers on request.
It runs on Elastic: in the Elastic you already run, a customer-owned Elastic Cloud deployment in the EU, or self-managed.
Keep it for commitments and rightsizing. We add AI providers and seats in one open model, cost next to usage, and a way to compare AI providers on value, inside your own Elastic.
No. We process billing data and usage metadata such as tokens, models and seats, not prompts or responses.
With a free one-hour scan, then a fixed-price baseline on your own bills.
Free scan