A FinOps data-to-action platform that transforms cloud cost signals into engineering actions and validated financial outcomes.
Cloud cost data is often fragmented across AWS CUR, Kubernetes allocation data, observability vendors, AI providers, and SaaS platforms. This makes it difficult to connect spend to ownership, product usage, forecasts, and optimization actions.
- Cost data ingestion
- Normalized cost schema
- Vendor and cloud spend categorization
- Ownership mapping
- Monthly cost summary
- Engineering-facing cost reporting
- Foundation for dashboards, alerts, and Jira workflows
- Python
- CSV-based prototype data
- AWS CUR-style cost model
- SQL examples
- Future: Athena, S3, Glue, Terraform, CI/CD, Jira API
Raw Cost Data → Normalize → Enrich with Ownership → Summarize → Report → Route Action
- Identify top cost drivers by service
- Map spend to team/application owner
- Create monthly FinOps review summaries
- Support forecasting and variance analysis
- Feed dashboards and Jira-based remediation workflows
This project shows how FinOps moves beyond reporting into operational cost governance: creating a repeatable system where cost signals become engineering decisions and business outcomes.
To transform fragmented cloud and vendor cost data into:
- Structured cost models
- Unit economics
- Ownership-driven insights
- Actionable workflows
Sources → Ingestion → Modeling → Storage → Action Layer
This platform reflects how I operationalize FinOps in practice, not just surfacing cost data, but translating it into ownership, action, and measurable outcomes.
It follows the same principles I’ve used in production:
- Cost signals must map to a clear owner
- Insights must translate into engineering actions
- Every action must be tracked and validated (e.g., via Jira workflows)
The goal is not visibility alone — it is execution, accountability, and continuous cost improvement.
AWS CUR ─┐
GCP ─────┼──> Ingestion Layer ──> Cost Modeling ──> Data Warehouse ──> Action Layer (Jira)
K8s ─────┤
SaaS ────┘
## 🧠 Architecture Overview
```text
┌───────────────┐
│ Cost Sources │
│───────────────│
│ AWS (CUR) │
│ GCP Billing │
│ Kubernetes │
│ SaaS Vendors │
│ AI Providers │
└───────┬───────┘
│
▼
┌───────────────┐
│ Ingestion │
│───────────────│
│ S3 + Athena │
│ APIs │
│ ETL Pipelines │
└───────┬───────┘
│
▼
┌───────────────┐
│ Cost Modeling │
│───────────────│
│ Unit Economics│
│ Cost Drivers │
│ Allocation │
└───────┬───────┘
│
▼
┌───────────────┐
│ Data Platform │
│───────────────│
│ Warehouse │
│ (Snowflake) │
└───────┬───────┘
│
▼
┌───────────────┐
│ Action Layer │
│───────────────│
│ Jira Tickets │
│ Ownership │
│ Alerts │
└───────────────┘
---
## 🔧 Core Capabilities
- Unified cost ingestion pipelines
- Cost driver and unit economics modeling
- Kubernetes cost allocation (shared + idle)
- Commitment strategy optimization (RIs, Savings Plans)
- Automated anomaly detection
- Jira-based action routing and ownership tracking
- AI-powered cost explanation layer
---
## 🧩 Platform Layers
### 1. Ingestion
Collects cost and usage data from multiple systems.
### 2. Modeling
Transforms raw data into cost drivers and business metrics.
### 3. Allocation
Distributes shared and platform costs across teams.
### 4. Action Layer
Routes insights into Jira for execution and tracking.
### 5. AI Layer
Explains anomalies and cost drivers in engineering terms.
---
## 🔗 Supporting Repositories
- reveal-finops-lab (cost ingestion + anomaly detection)
- driver-based-finops-modeling-engine (unit economics)
- Kubernetes-Economic-Attribution-Engine (allocation)
- claude-lambda-finops-agent (AI reasoning layer)
---
## 🚀 Outcome
This platform enables organizations to:
- Move from cost visibility → cost action
- Align engineering and finance
- Validate savings in actual billing data
- Build a scalable FinOps operating model
---
---
## 📊 Key FinOps Metrics
This platform is designed to support decision-making through measurable outcomes:
- Commitment Coverage %
- Commitment Utilization %
- Cost per Unit (API, user, transaction)
- Idle Resource %
- Anomaly Detection & Resolution Time
- Budget vs Forecast Variance
These metrics connect infrastructure cost directly to business performance and engineering behavior.
---
## 🔗 Supporting Components
This platform is composed of modular components, each focused on a specific layer of the FinOps system:
- **Cost Ingestion** → [reveal-finops-lab](https://github.com/drchristinee1/reveal-finops-lab)
Pipelines for ingesting AWS CUR data and generating cost signals
- **Cost Modeling** → [driver-based-finops-modeling-engine](https://github.com/drchristinee1/driver-based-finops-modeling-engine)
Unit economics and cost driver modeling engine
- **Kubernetes Allocation** → [Kubernetes-Economic-Attribution-Engine](https://github.com/drchristinee1/Kubernetes-Economic-Attribution-Engine)
Allocation of shared and idle Kubernetes costs across teams
- **AI Cost Analysis** → [claude-lambda-finops-agent](https://github.com/drchristinee1/claude-lambda-finops-agent)
AI-powered explanation of cost anomalies and engineering recommendations
## 🧪 Examples
This repository includes sample artifacts demonstrating how the platform operates:
- **AWS CUR Query**
Example SQL used to aggregate cost data:
`examples/sample_cur_query.sql`
- **Anomaly → Jira Automation**
Python example that converts cost anomalies into Jira tickets:
`examples/anomaly_to_jira.py`
## Real-World Application
This architecture is inspired by the FinOps operating model I’ve built, where cost variance signals are automatically routed to technical owners through Jira workflows, tracked through resolution, and used to improve forecasting and accountability.
This approach defines cost-efficient patterns across compute, storage, observability, and AI usage, and embeds them into infrastructure and workflows so cost efficiency becomes a built-in part of engineering decisions, not an afterthought.