How SkillGigs Cut AWS Costs by 92% Without Performance Loss

From $87.5K to $7K monthly spend in 6 months

Challenge

This story hits home for almost every growing tech company that moved to the cloud “to save money” and then got a nasty surprise.

Problem

Value of optimization

Before

They likely ran a traditional, monolithic application on always-on EC2 servers (e.g., a large web server, a large application server, a large database server). These servers were running 24/7, consuming resources (and costing money) even during periods of low traffic (nights, weekends).

After

They broke their application into smaller, independent functions (microservices) and ran them on AWS Lambda.

Saving

With Lambda, you pay only for the milliseconds your code executes. There are no charges when code is not running. This eliminated the cost of idle compute resources, which is the biggest source of waste in traditional architectures.

Pain Points That Resonate with Similar Companies

This story hits home for almost every growing tech company that moved to the cloud “to save money” and then got a nasty surprise.

Pain Points

Our AWS Bill is Unpredictable and Scary

We're Throwing Money at Idle Resources

Our Team Spends Too Much Time on DevOps, Not Our Product

We're Stuck in a Monolithic Architecture

The Cloud Cost Conversation is a Black Box

The bill fluctuates wildly month-to-month, making financial forecasting a nightmare. CEOs and CFOs lose sleep.

Development, staging, and test environments run 24/7 but are used only 40 hours a week. Production servers are over-provisioned "just to be safe."

Engineers are busy patching OSes, scaling servers, and managing database clusters instead of building new features.

The application is a giant, intertwined codebase. It's hard to scale, update, and innovate. Scaling one feature means scaling the entire expensive monolith.

Developers who build features don't see the cost implications of their choices. The bill goes to a separate finance department, creating a disconnect.

Cost Optimization With EC2 Spot Instances

Containerization with ECS

Spot Fleet

Instance Diversification

Cost
Optimization

Solution

Start with
Easy Wins

Cost Optimization
is a Cycle

Measure
& Monitor

Optimize for
Efficiency

Step by Step Approach

Cost Optimization is a Cycle, Not a Project:

It's continuous. Establish a culture of cost awareness.

Start with Easy Wins:

Build momentum with quick savings to justify more complex efforts.

Measure & Monitor: 

You can't optimize what you can't measure. Use AWS's native tools extensively.

Optimize for Efficiency, Not Just Cost:

The goal is to get the most value out of every dollar spent, not necessarily to spend the least. Cutting costs in a way that hurts performance or reliability is a failure.

Timeline of implementations

Build and test Amazon Server

Using Terraform develop code

Deploy Code

Code will be deployed by a different team

5 DaysSet up network firewall

Network team will work on defining and deploying the firewall

Build VPC

Build VPC The VPC needs to be defined before servers can be built

Peer-review Code

Paperwork submitted for code has to be reviewed and approved by a lead or pee

Build and test Amazon Server

Active directory team will do the join using ticketing process

Case Study

By submitting this form, you agree to our privacy policy. Your information will never be shared.

Case Study

By submitting this form, you agree to our privacy policy. Your information will never be shared.

Case Study

By submitting this form, you agree to our privacy policy. Your information will never be shared.
Your submission was successful.
Sign up to continue

By signing up, I accept the Cloudlogically Terms of Service and acknowledge the Privacy Policy.

Or continue with:
[social-login provider='google']