FinOps Learning Path 2026: Roadmap, FOCUS & FinOps for AI

What is FinOps?

FinOps is an operational framework and cultural practice that maximises the business value of technology, enables timely data-driven decisions and creates financial accountability through collaboration between engineering, finance and business teams. It is not a cost-cutting project. The goal is to spend the right amount on the right things and to know what each unit of spend returns.

The discipline has broadened fast. The 2026 FinOps Framework update added an Executive Strategy Alignment capability and dedicated guidance for five technology categories: public cloud, SaaS, data centre, data cloud platforms and AI. The State of FinOps 2026 survey found that 98 percent of practitioners now manage AI spend and 90 percent manage SaaS, and FOCUS 1.4, ratified in June 2026, together with native FOCUS exports from AWS, Microsoft and Google Cloud, finally makes cost data from different providers comparable.

The path follows the FinOps lifecycle. You start by building trustworthy visibility and allocation (Inform), move to rate, usage and Kubernetes optimisation (Optimize), and finish by running the practice: forecasting, governance, sustainability, automation and FinOps for AI and the wider Cloud+ scope (Operate). Every stage is hands-on with real billing exports, SQL, OpenCost, policy-as-code and IaC cost estimation, alongside the communication skills a FinOps practitioner needs.

FinOps works closely with CloudOps and platform engineering, who build the guardrails and automation, with DevOps for shift-left cost checks in pipelines, and with MLOps and LLMOps on GPU and token costs. It also intersects IT asset management, IT financial management, ITSM and sustainability teams.

What’s new in 2026

The 2026 FinOps Framework update added the Executive Strategy Alignment capability and dedicated guidance for Public Cloud, SaaS, Data Center, Data Cloud Platforms and AI, completing the shift from cloud cost to technology value.
AI is now core FinOps work: the State of FinOps 2026 found 98 percent of respondents manage AI spend, up from 31 percent two years earlier, and named FinOps for AI the top forward-looking priority.
FOCUS 1.4 was ratified in June 2026, adding Invoice Detail and Billing Period datasets and much richer contract commitment data, while AWS, Microsoft and Google Cloud all offer native FOCUS exports.
Cost decisions are shifting left: pre-deployment architecture costing is one of the most requested FinOps capabilities, pushing cost estimates into design reviews, IaC and pull requests.
Commitment models keep getting more flexible, such as AWS Database Savings Plans launched in December 2025, making automated, portfolio-style commitment management the norm.
FinOps teams increasingly sit in technology leadership, with 78 percent now reporting to the CTO or CIO, and are adopting AI assistants and MCP servers that query billing data directly.

Who this path is for

Cloud, DevOps and platform engineers who want to design and run cost-efficient infrastructure
Finance, procurement and IT financial management professionals moving into cloud and AI spend
Data and BI analysts who build cost reporting, forecasting and unit-economics dashboards
Engineering managers and product owners accountable for what their teams spend
IT asset and SaaS management professionals expanding into FinOps beyond public cloud

Before you start

Basic familiarity with at least one public cloud (AWS, Azure or Google Cloud) and its core services
Comfort with spreadsheets and basic SQL
Basic business finance vocabulary such as budgets, opex vs capex and amortisation (taught in context if new)
For the Kubernetes and automation stages: working knowledge of containers, Kubernetes and Terraform or OpenTofu

What you will be able to do

Explain and apply the 2026 FinOps Framework: principles, phases, domains, capabilities, personas, scopes and maturity
Ingest and normalise AWS, Azure and Google Cloud billing data with the FOCUS specification and analyse it with SQL
Design tagging, hierarchy and allocation strategies that make over 90 percent of spend allocatable, with showback and chargeback
Define unit economics and KPIs that connect technology spend to business value
Plan commitment purchases and run a rightsizing and waste-removal programme with quantified, tracked savings
Allocate and optimise Kubernetes costs with OpenCost, rightsizing and autoscaling without hurting reliability
Measure, allocate and optimise AI, GPU, SaaS and licensing spend alongside public cloud
Run an operating FinOps practice with forecasting, budgets, anomaly management and policy-as-code guardrails

The roadmap

6 stages · 23 topics · ~192 hours of focused learning. Work through the stages in order; optional topics are electives you can come back to.

1
Beginner 3 weeks ~24 h 3 topics

Stage 1 · FinOps Foundations

What FinOps is in 2026, how the FinOps Foundation framework is structured, how cloud pricing and billing really work, and how a FinOps practice is organised.
Mark “The FinOps Framework (2026)” as done
The FinOps Framework (2026)
10h

Learn the six principles, the Inform, Optimize and Operate phases, the four domains and their capabilities, personas, scopes and the crawl, walk, run maturity model. The 2026 update adds Executive Strategy Alignment and technology categories, so FinOps now covers AI, SaaS, data cloud platforms and data centres as well as public cloud.
Key concepts & resources

Key concepts

six FinOps principles
Inform, Optimize and Operate phases
four domains and their capabilities
core and allied personas
scopes and technology categories
crawl, walk, run maturity

Learn it from

FinOps Framework (FinOps Foundation) Docs

Introduction to FinOps (free course) Course

FinOps Training — DevOpsSchool Course Paid
Mark “Cloud Pricing & Billing Fundamentals” as done
Cloud Pricing & Billing Fundamentals
8h

Understand how providers charge: on-demand, commitment discounts, spot, tiered and data-transfer pricing, credits and marketplace purchases. Learn the cost measures that confuse everyone at first, such as list, billed, effective and amortised cost, because every report you build depends on them.
Key concepts & resources

Key concepts

on-demand, commitment and spot pricing
data transfer and egress charges
list vs billed vs effective cost
amortisation of upfront fees
credits, refunds and marketplace charges
invoice data vs usage data

Learn it from

AWS Well-Architected Framework: Cost Optimization Pillar Docs

Microsoft FinOps Documentation Docs

Cloud Training — DevOpsSchool Course Paid
Mark “FinOps Culture, Personas & Operating Model” as done
FinOps Culture, Personas & Operating Model
6h

Set up the people side: a central team that enables rather than polices, clear responsibilities for engineering, finance, procurement, product and leadership, and a regular cadence of reviews. Most FinOps failures are organisational, not technical.
Key concepts & resources

Key concepts

centrally enabled, federated ownership
responsibilities across personas
executive sponsorship
team placement and reporting lines
review cadences and communication
intersecting disciplines: ITAM, ITFM and ITSM

Learn it from

FinOps Personas Docs

FinOps Practice Operations Capability Docs

Cloud FinOps, 2nd Edition (O'Reilly) Book Paid
Milestone: Explain the FinOps Framework to engineering and finance colleagues and map your organisation's personas, scopes and current maturity.
2
Beginner 4 weeks ~32 h 4 topics

Stage 2 · Cost Visibility & Billing Data

The Inform phase: getting trustworthy cost and usage data into one place through native billing exports, the FOCUS specification and solid data modelling, then giving each persona the reports and anomaly alerts they need.
Mark “Native Billing Data on AWS, Azure & Google Cloud” as done
Native Billing Data on AWS, Azure & Google Cloud
8h

Set up AWS Data Exports (CUR 2.0 and FOCUS), Azure Cost Management exports and the Google Cloud Billing export to BigQuery, and learn each provider's quirks. Detailed line-item billing data is the raw material for every later stage.
Key concepts & resources

Key concepts

AWS Data Exports and CUR 2.0
Azure Cost Management exports
Google Cloud Billing export to BigQuery
billing hierarchies: organisations, accounts, subscriptions and projects
data latency and restatements
price sheets and pricing data

Learn it from

AWS Data Exports User Guide Docs

Microsoft Cost Management and Billing Documentation Docs

Export Cloud Billing Data to BigQuery (Google Cloud) Docs
Mark “The FOCUS Specification” as done
The FOCUS Specification
8h

Learn FOCUS, the FinOps Open Cost and Usage Specification that normalises billing data across clouds, SaaS and other providers. Version 1.4 (June 2026) adds Invoice Detail and Billing Period datasets and expands contract commitment data, so invoices, usage and commitments can be reconciled in one schema.
- DuckDB
Key concepts & resources

Key concepts

FOCUS datasets and columns
BilledCost, EffectiveCost, ListCost and ContractedCost
ChargeCategory and commitment discount columns
provider-specific x_ columns
conformance and version differences
invoice reconciliation

Learn it from

FOCUS Specification (current version) Docs

Introduction to FOCUS (free course) Course

FOCUS Specification Repository on GitHub Docs
Mark “Cost Data Ingestion, Modelling & Analytics” as done
Cost Data Ingestion, Modelling & Analytics
10h

Build a cost data platform: land exports in object storage, model them with SQL, join organisational and business metadata, and serve self-service dashboards. Use DuckDB for fast exploration and Microsoft's FinOps toolkit and the AWS Cloud Intelligence Dashboards as reference implementations.
- DuckDB
- Python
- Power BI
Key concepts & resources

Key concepts

landing billing data in object storage
SQL modelling of cost data
joining tags, org and business metadata
multi-cloud normalisation
freshness and completeness checks
self-service BI

Learn it from

Microsoft FinOps Toolkit Tool

AWS Cloud Intelligence Dashboards Workshop Lab

Power BI Training — DevOpsSchool Course Paid
Mark “Reporting, Showback & Anomaly Management” as done
Reporting, Showback & Anomaly Management
6h

Design reports for each persona, from engineering team views to executive summaries, and set up cost anomaly detection with a clear triage and ownership workflow. An unexpected spike found within hours costs far less than one found on the monthly invoice.
- AWS
- Microsoft Azure
Key concepts & resources

Key concepts

persona-specific dashboards
showback reports
cost anomaly detection
anomaly triage and ownership
alert thresholds and noise
variance commentary

Learn it from

FinOps Anomaly Management Capability Docs

AWS Cost Anomaly Detection Docs

Monitor Usage and Spending with Cost Alerts (Azure) Docs
Milestone: Load FOCUS-formatted billing data from at least two providers into one queryable dataset and publish persona-specific dashboards with anomaly alerts.
3
Intermediate 4 weeks ~32 h 4 topics

Stage 3 · Allocation, Accountability & Unit Economics

Making every cost someone's cost: tagging and hierarchy strategy, shared-cost allocation, showback and chargeback, unit economics and the automated guardrails that keep allocation accurate over time.
Mark “Tagging & Account Hierarchy Strategy” as done
Tagging & Account Hierarchy Strategy
8h

Design an account, subscription and project hierarchy plus a small mandatory tag set such as owner, cost centre, application and environment. Learn how each provider activates tags for billing and how to measure and improve tag coverage.
Key concepts & resources

Key concepts

account, subscription and project hierarchy
mandatory vs optional tags
cost allocation tag activation
tag coverage KPI
handling untaggable resources
tagging in IaC modules

Learn it from

Best Practices for Tagging AWS Resources Paper

Define Your Tagging Strategy (Azure Cloud Adoption Framework) Docs

Google Cloud Tags Overview Docs
Mark “Shared Costs, Showback & Chargeback” as done
Shared Costs, Showback & Chargeback
8h

Split shared platform, support, networking and commitment costs fairly using even, proportional or fixed methods, and reconcile allocations with the invoice. Decide when showback is enough and when real chargeback to budgets is worth the effort.
- Microsoft Azure
Key concepts & resources

Key concepts

direct vs shared costs
even, proportional and fixed splits
amortising commitments and support fees
showback vs chargeback
invoice reconciliation
allocation rules as code

Learn it from

FinOps Allocation Capability Docs

FinOps Invoicing & Chargeback Capability Docs

Introduction to Cost Allocation (Azure) Docs
Mark “Unit Economics & FinOps KPIs” as done
Unit Economics & FinOps KPIs
8h

Connect spend to value with unit costs such as cost per customer, per transaction or per thousand tokens, and track KPIs like allocatable spend, effective savings rate and forecast accuracy. Unit economics lets you welcome rising spend when it means a growing business.
- Python
Key concepts & resources

Key concepts

unit cost metrics
business driver data
marginal vs average cost
KPI definitions and targets
benchmarking
value realisation reporting

Learn it from

FinOps Unit Economics Capability Docs

FinOps KPIs & Benchmarking Capability Docs
Mark “Governance & Policy as Code” as done
Governance & Policy as Code
8h

Encode cost policies, such as required tags, allowed instance types, idle-resource clean-up and budget gates, with Cloud Custodian and OPA, and apply them in IaC pipelines. Automated guardrails scale where emails and spreadsheets do not.
Key concepts & resources

Key concepts

Governance, Policy & Risk capability
preventive vs detective controls
Cloud Custodian policies
OPA checks on Terraform plans
exceptions and waivers
automated clean-up of waste

Learn it from

Cloud Custodian Documentation Docs

FinOps Governance, Policy & Risk Capability Docs

Terraform Training — DevOpsSchool Course Paid
Milestone: Reach and report over 90 percent allocatable spend under an enforced tagging policy, and publish at least two unit-cost metrics tied to business drivers.
4
Intermediate 5 weeks ~40 h 4 topics

Stage 4 · Rate & Usage Optimisation

The Optimize phase: paying less per unit through commitment discounts, using fewer units through rightsizing and waste removal, designing for cost before deployment, and taming data platform spend.
Mark “Rate Optimisation: Commitments & Discounts” as done
Rate Optimisation: Commitments & Discounts
12h

Plan and manage AWS Savings Plans and Reserved Instances, Azure reservations and savings plans, and Google Cloud committed use discounts. Learn coverage, utilisation and effective savings rate, and how to ladder commitments so savings stay high as usage changes.
Key concepts & resources

Key concepts

Savings Plans, reservations and CUDs
coverage and utilisation
effective savings rate
commitment laddering and term mix
spot and preemptible capacity
enterprise agreements and negotiated discounts

Learn it from

What Are Savings Plans? (AWS) Docs

Azure Savings Plan Overview Docs

Committed Use Discounts (Google Cloud) Docs
Mark “Usage Optimisation & Rightsizing” as done
Usage Optimisation & Rightsizing
12h

Find and remove waste: idle and orphaned resources, oversized instances and databases, unattached storage, avoidable data transfer and non-production environments running all weekend. Start from provider recommendation engines, then validate against real utilisation before acting.
Key concepts & resources

Key concepts

rightsizing compute and databases
idle and orphaned resources
scheduling non-production
storage tiering and lifecycle policies
data transfer reduction
newer instance generations and ARM

Learn it from

AWS Cost Optimization Hub Docs

Azure Advisor Cost Recommendations Docs

Google Cloud Active Assist and Recommender Docs
Mark “Cost-Aware Architecture & Shift-Left” as done
Cost-Aware Architecture & Shift-Left
10h

Make cost a design input: compare serverless, containers and VMs, choose regions and data paths with egress in mind, and show the cost impact of infrastructure changes in pull requests with Infracost. The cheapest resource is the one you never deploy.
Key concepts & resources

Key concepts

Architecting & Workload Placement capability
cost trade-offs between compute models
egress-aware design
cost estimates in pull requests
planning and estimating new workloads
cost as a non-functional requirement

Learn it from

Infracost Documentation Docs

FinOps Architecting & Workload Placement Capability Docs

OpenTofu Training — DevOpsSchool Course Paid
Mark “Data Cloud Platform Costs” as done
Data Cloud Platform Costs
Optional 6h

Apply FinOps to Snowflake, Databricks and BigQuery, where credits, DBUs and slots behave very differently from VMs. Data cloud platforms are now a named FinOps technology category and often one of the largest line items.
Key concepts & resources

Key concepts

Snowflake credits and warehouses
Databricks DBUs and billing system tables
BigQuery on-demand vs capacity pricing
attributing query and job costs
auto-suspend and right-sized clusters

Learn it from

FinOps Technology Category: Data Cloud Platforms Docs

Understanding Overall Cost (Snowflake) Docs

Billable Usage System Table Reference (Databricks) Docs
Milestone: Deliver a prioritised optimisation backlog with quantified savings and a commitment plan with target coverage, utilisation and effective savings rate.
5
Advanced 3 weeks ~24 h 3 topics

Stage 5 · Kubernetes & Container Cost Management

Allocating and optimising shared clusters, where the cloud bill shows nodes but teams own namespaces and workloads.
Mark “Kubernetes Cost Allocation with OpenCost” as done
Kubernetes Cost Allocation with OpenCost
10h

Deploy OpenCost, the CNCF incubating project for Kubernetes cost allocation, to split cluster spend by namespace, label and workload using requests and actual usage. Learn how idle capacity and shared system costs are handled and how provider data such as AWS split cost allocation complements it.
Key concepts & resources

Key concepts

requests vs usage-based allocation
idle and unallocated capacity
shared system and platform costs
namespace and label allocation
OpenCost specification
cloud billing integration

Learn it from

OpenCost Documentation Docs

OpenCost Specification Docs

FinOps for Containers (FinOps Foundation course) Course Paid
Mark “Commercial Kubernetes Cost Platforms” as done
Commercial Kubernetes Cost Platforms
Optional 4h

Evaluate IBM Kubecost, now part of IBM's Apptio portfolio, and similar tools for multi-cluster views, savings insights, network costs and budgets. Know exactly what they add on top of OpenCost so you can justify, or avoid, the licence.
- Kubecost
Key concepts & resources

Key concepts

multi-cluster cost aggregation
savings recommendations
network and storage cost
alerts and budgets
build vs buy

Learn it from

IBM Kubecost Tool

IBM Kubecost 3.0: Faster, Smarter, and Built for Scale Article
Mark “Rightsizing & Autoscaling for Cost” as done
Rightsizing & Autoscaling for Cost
10h

Set requests and limits from real usage, apply the Vertical and Horizontal Pod Autoscalers, and let Karpenter consolidate nodes and use spot capacity. Done well this is the biggest Kubernetes saving; done badly it causes throttling and outages, so measure SLOs alongside cost.
- Kubernetes
- Karpenter
Key concepts & resources

Key concepts

requests and limits from usage
Vertical Pod Autoscaler
Horizontal Pod Autoscaler
node consolidation and bin packing
spot node pools and interruption handling
measuring savings against SLOs

Learn it from

Resource Management for Pods and Containers (Kubernetes) Docs

Vertical Pod Autoscaling (Kubernetes) Docs

Karpenter Documentation Docs
Milestone: Report per-team Kubernetes cost including idle and shared overhead, and reduce cluster cost through rightsizing and autoscaling with no SLO regressions.
6
Advanced 5 weeks ~40 h 5 topics

Stage 6 · FinOps for AI, Cloud+ & Operating the Practice

The Operate phase and the wider scope: AI and GPU spend, SaaS, licensing and data centres, plus forecasting, budgeting, sustainability and the automation that keeps a FinOps practice running.
Mark “FinOps for AI & GPU Spend” as done
FinOps for AI & GPU Spend
10h

Track and allocate AI costs across token-based APIs, managed AI platforms and self-hosted GPUs, and measure cost per inference or per business outcome. Learn the levers that matter: model choice, prompt and context size, caching, batching, provisioned versus on-demand capacity and GPU utilisation.
Key concepts & resources

Key concepts

token economics and pricing tiers
cost per inference and per outcome
provisioned throughput vs on-demand
GPU utilisation and sharing
attributing AI spend to products
measuring AI value

Learn it from

FinOps for AI Overview (FinOps Foundation) Paper

OpenTelemetry Semantic Conventions for Generative AI Docs

NVIDIA DCGM Exporter Docs
Mark “SaaS, Licensing & Data Center Spend” as done
SaaS, Licensing & Data Center Spend
6h

Apply FinOps capabilities to SaaS subscriptions, software licences and on-premises data centres, working closely with IT asset management and procurement. Most organisations now manage these alongside cloud, and FOCUS increasingly covers SaaS billing too.
Key concepts & resources

Key concepts

SaaS discovery and seat utilisation
licence optimisation and bring-your-own-licence
renewal and contract timing
data centre cost models
working with ITAM and procurement

Learn it from

FinOps Licensing & SaaS Capability Docs

FinOps Technology Category: SaaS Docs

FinOps Technology Category: Data Center Docs
Mark “Forecasting, Budgeting & Planning” as done
Forecasting, Budgeting & Planning
10h

Forecast spend from trends, seasonality and business drivers, set budgets with named owners and alerts, and estimate new workloads before launch. Track forecast accuracy and connect major investments to executive strategy and multi-year plans.
- Python
Key concepts & resources

Key concepts

trend and driver-based forecasting
forecast accuracy and variance
budget ownership and alerts
planning and estimating new workloads
multi-year investment planning
Executive Strategy Alignment capability

Learn it from

FinOps Forecasting Capability Docs

FinOps Budgeting Capability Docs

FinOps Executive Strategy Alignment Capability Docs
Mark “Sustainability & GreenOps” as done
Sustainability & GreenOps
6h

Measure and reduce the carbon footprint of technology using provider carbon tools, Kubernetes energy metrics and the Software Carbon Intensity specification. Many carbon reductions are also cost reductions, which makes sustainability a natural FinOps partner.
- Kepler
- AWS
- Google Cloud
Key concepts & resources

Key concepts

Scope 1, 2 and 3 emissions
provider carbon footprint tools
Software Carbon Intensity (SCI)
Kubernetes energy measurement
carbon-aware region and time shifting

Learn it from

FinOps Sustainability Capability Docs

Software Carbon Intensity (SCI) Specification Paper

Kepler: Kubernetes Efficient Power Level Exporter Tool
Mark “Automation, AI Assistants & Practice Operations” as done
Automation, AI Assistants & Practice Operations
8h

Automate the repetitive parts of FinOps, from anomaly-to-ticket workflows to scheduled clean-ups and commitment reports, and use AI assistants or MCP servers to query billing data with read-only access. Assess maturity regularly and run the practice with clear cadences and ongoing education.
- Python
- AWS
Key concepts & resources

Key concepts

Automation, Tools & Services capability
anomaly-to-ticket workflows
MCP servers for billing data
read-only access and guardrails for AI assistants
FinOps assessment and maturity
education and enablement

Learn it from

AWS Billing and Cost Management MCP Server Tool

FinOps Automation, Tools & Services Capability Docs

FinOps Assessment Capability Docs
Milestone: Run a monthly FinOps cadence covering cloud, AI and SaaS spend, with forecasts inside agreed variance, executive reporting and automated anomaly-to-action workflows.
You made it. Keep sharpening with the projects and certifications below.

The tool stack

21 tools and platforms used across this path. Focus on concepts first — tools change, principles don’t.

Cloud Platforms & Governance

Containers & Kubernetes

Data Engineering

FinOps & Sustainability

Foundations & Languages

Python

Infrastructure as Code & Automation

Observability

Policy as Code & Secrets

Portfolio projects

Build these to prove your skills. Publish each one on GitHub with a clear README and architecture diagram.

01 Beginner

Multi-Cloud FOCUS Cost Lake

Load FOCUS exports, or FOCUS sample data, from two providers into DuckDB, build SQL models for cost by service, team and charge category, and publish a dashboard. Document every metric definition, including billed versus effective cost.

FOCUS
SQL
DuckDB
cost reporting

02 Intermediate

Tagging & Showback Programme

Define a tag policy, enforce it with Cloud Custodian or OPA in the IaC pipeline, and build a tag-coverage report. Allocate shared platform costs proportionally and deliver a monthly showback pack for three teams.

tagging strategy
policy as code
shared cost allocation
showback

03 Intermediate

Commitment & Rightsizing Plan

Analyse 90 days of usage to recommend a laddered Savings Plan, reservation or CUD purchase with target coverage and effective savings rate, plus a prioritised rightsizing backlog. Present savings, risk and a review cadence as you would to a CFO.

rate optimisation
rightsizing
financial analysis
stakeholder communication

04 Advanced

Kubernetes Cost Allocation & Optimisation

Deploy OpenCost on a cluster running several team workloads and report per-namespace cost including idle capacity, then apply VPA recommendations and Karpenter consolidation. Show before-and-after cost alongside latency and error-rate SLOs.

OpenCost
Kubernetes rightsizing
autoscaling
cost allocation

05 Advanced

AI Unit-Economics Dashboard

Instrument an LLM application with OpenTelemetry GenAI attributes, join token usage with provider billing data, and report cost per request, per customer and per successful outcome. Model the savings from caching, smaller models and batching.

FinOps for AI
unit economics
OpenTelemetry
forecasting

06 Advanced

Shift-Left Cost Guardrails & Anomaly Automation

Add Infracost estimates and a budget policy check to pull requests, and route cost anomaly alerts into tickets with owner assignment. Include a read-only assistant that answers cost questions from your FOCUS dataset through an MCP server.

Infracost
CI/CD integration
anomaly management
FinOps automation

Certifications worth earning

Certifications that validate this path’s skills. Always confirm the current exam version on the official page before booking.

Foundation

FinOps Foundation Certified Practitioner (FinOps-FCP)

DevOpsSchool

DevOpsSchool's instructor-led, project-based FinOps programme aligned to this path; it is a DevOpsSchool credential, separate from the FinOps Foundation's own FOCP exam.

Official page

Foundational