For many research institutions, moving workloads to AWS has unlocked the scale and flexibility that modern scientific computing requires.
But there is one area where the experience remains consistently difficult:
Managing and accounting for cloud costs.
Research IT teams, Principal Investigators, and finance departments frequently struggle to answer questions like:
- Which project is consuming the budget?
- Are we charging the correct Cost Centre?
- What resources are running across all AWS accounts?
- Which Slurm jobs drove this month’s GPU costs?
Most cloud cost management tools are designed for enterprise IT environments—not for grant-funded, multi-project research ecosystems.
The result?
Manual spreadsheets. Tagging gaps. Missing assets. Billing surprises.
FinOps for research computing has become unnecessarily complex and operationally frustrating.
To solve this, Relevance Lab worked closely with AWS teams and research institutions to understand real operational challenges.
Announcing Research Gateway FinOps — a purpose-built FinOps solution designed specifically for research environments.
The Top Three FinOps Needs from Research Customers
1. Project-Based Tracking and Accounting
In research environments, funding is allocated to projects, not AWS accounts.
Institutions need financial visibility across a hierarchy that looks like:
Cost Centre → Project → User → Job
Typical scenarios include:
- Projects linked to AWS accounts
- One AWS account may support multiple projects
- Projects mapped to Cost Centres with specific cost codes
- Multiple projects may belong to the same Cost Centre
Research Gateway FinOps enables institutions to manage this complexity with ease by providing:
- Monthly project-level consumption views
- Budget vs. actual spend tracking
- Drill-down visibility to users, jobs, and services
- Finance-ready monthly chargebacks
This ensures financial reporting aligns with how research funding is actually structured.
Project View
Account View
Cost Code View
2. Complete Visibility of All Assets Across AWS Accounts
One of the most common problems in research cloud environments is simply understanding:
What resources are running — and where.
Typical challenges include:
- Resources created without proper tags
- Legacy environments with inconsistent tagging standards
- Shared infrastructure costs across projects
- Orphaned or unused resources
Without full visibility into cloud assets, accurate chargebacks become nearly impossible.
What if you could start with your existing AWS environment?
Research Gateway FinOps is designed to work with your existing AWS environment.
Day 1: Automated Discovery
The platform:
- Scans all onboarded AWS accounts
- Discovers every active resource
- Identifies untagged or incorrectly tagged assets
- Maps existing tags where available
Through a self-service portal, administrators can:
- Assign resources to Projects and Cost Centres
- Apply tagging corrections in bulk
- Establish governance policies for future resources
Ongoing Governance: Phantom Asset Detection
After initial cleanup, the platform continuously monitors your environment.
Research Gateway FinOps:
- Continuously scans AWS accounts
- Detects new untagged or phantom assets
- Highlights them in daily reports
- Allows administrators to map them to the correct Project directly from the UI
This ensures:
- No cost remains unaccounted
- New resources stay within governance controls
- Chargebacks remain accurate month after month
Phantom Assets Detection View with GenAIAgent
This “discover, clean, and continuously watch” model makes FinOps frictionless for research teams.
3. Deep HPC and Slurm Cost Tracking for High-Cost Resources
For many research institutions, the largest cloud expenditures come from:
- GPU clusters
- Large-scale compute jobs
- Elastic HPC environments
These workloads are commonly managed through:
- AWS ParallelCluster
- Slurm
- Open OnDemand
However, AWS billing alone cannot answer critical operational questions:
- Which jobs ran?
- Which users consumed GPU hours?
- How does cluster usage map to project budgets?
Research Gateway FinOps integrates multiple operational data sources including:
- Slurm accounting data
- Open OnDemand usage
- AWS Cost & Usage Reports
The platform converts operational workload data into accurate project-level cloud cost attribution, particularly for high-value GPU workloads.
This allows research institutions to connect compute activity directly to funding consumption.
Built for Existing Environments, Not Just New Ones
Many FinOps initiatives fail because they assume a perfectly tagged and well-governed environment from the start.
Research Gateway FinOps takes a different approach:
- Works with existing AWS accounts
- Automatically discovers current resources
- Enables self-service cleanup and governance
- Continuously enforces cost accountability
Institutions can move from limited visibility to full research FinOps in days instead of months.
A Simpler Way to Manage Research Cloud Costs
Research Gateway FinOps brings together everything research organizations need in a single platform:
- Project-based financial visibility
- Full asset discovery and governance
- Continuous phantom asset monitoring
- HPC and Slurm workload cost attribution
- Monthly Cost Centre chargebacks
- Budget tracking and guardrails
Instead of relying on manual reconciliation and reactive cost management, research teams gain a unified platform designed around how research funding and computing actually work.