Cloud Engineering
Serverless FinOps Optimization Suite
Coursework project for Cloud Economics (Sheridan College) — a Lambda cost analysis dashboard covering memory right-sizing, provisioned-concurrency optimization, and containerization recommendations across a simulated set of serverless functions. The analysis techniques are real; the function inventory, savings estimates, and forecasting numbers come from the course's sample dataset, not a live production workload.
Timeline
Fall 2025
Role
Student Developer
Team
1 person
Client
Coursework — Cloud Economics, Sheridan College
The Problem
Serverless costs are easy to lose track of — over-provisioned memory and idle provisioned concurrency quietly inflate the bill, and it's not obvious which functions are worth right-sizing versus containerizing. The exercise worked through a sample set of ~95 Lambda functions to practice identifying that waste.
The Solution
Built a dashboard that analyzes memory allocation against actual usage, evaluates provisioned-concurrency cost/benefit, flags containerization candidates, and forecasts potential savings — including a 3D visualization for comparing migration candidates.
Impact
- ▸$750–1,500/month in potential savings modeled (a 30–60% reduction) across the sample function set
- ▸~95 Lambda functions analyzed for memory right-sizing and provisioned-concurrency optimization
- ▸ML-based cost forecasting exercise showing 85–95% accuracy on the sample data
ValueCoursework demonstrating the serverless cost-optimization workflow — memory right-sizing, concurrency tuning, and containerization triage — used in real FinOps practice; the dollar figures describe the course dataset, not production Lambda spend.
Highlights
- ▸Lambda memory right-sizing and provisioned-concurrency cost analysis
- ▸Containerization recommendations for high-cost functions
- ▸ML-based cost forecasting with 3D visualization of migration candidates
Tech Stack