Cloud Cost Optimization: The GCP Audit That Paid for the Project
Cloud cost optimization services cut ElephantCPA's GCP infrastructure bill by 80%, an architecture overhaul, not a scope cut, that ran alongside 130+ shipped user stories and a tightened security posture for the accounting SaaS platform. The audit effectively paid for the rest of the project
- Geography
- Wayne, New Jersey, United States
- Stage
- Established SMB
- Team
- 1 senior + 1 mid full-stack engineer, 1 product designer, 1 product owner
The situation
ElephantCPA already had a working MVP that streamlined how accountants and their clients communicate. The platform was live and in production, but the next phase needed real depth: a substantially expanded feature set, an architecture that could carry more weight than the original MVP was built for, and a security posture that matched the sensitivity of the financial data running through it.
The brief was a platform enhancement, not a rewrite. The work spanned product, infrastructure, and security, with a parallel ask that made this a candidate for cloud cost optimization services from the start: take a hard look at the GCP cloud bill and find what was being over-provisioned. The two fronts weren't meant to trade off against each other, the goal was full product depth and a leaner cloud bill, without sacrificing one for the other.
What we built
The engagement was scoped as a fixed-scope, milestone-billed build, with a senior full-stack engineer leading and a mid full-stack engineer carrying the feature work alongside a product designer and a product owner.
On the product side, the team shipped 130+ user stories covering every meaningful surface of the platform: account management, document handling, accountant-client messaging, scheduling, and reporting. This kind of accounting software development work moved in paired design and engineering reviews, story by story, rather than as one large release.
On the infrastructure side, a Cloud Advisory review of the GCP setup surfaced significant over-provisioning. The recommendations, applied across compute and storage, reduced cloud spend by 80%. That reduction came from an architecture overhaul, not from narrowing the product roadmap — the same 130+ user stories shipped on the product side ran in parallel, untouched by the cost work. The same overhaul also tightened the platform's cybersecurity posture and stabilized the system end to end, so the savings did not come at the expense of reliability.
Tech stack: React with TypeScript on the frontend, Node.js on the backend, MySQL for storage, GCP Cloud Run for compute, with integrations to Stripe for billing, SendGrid for email, and Zoho for CRM. Components were built to be scalable so later feature work would not require re-architecting the platform again.
Outcome
ElephantCPA came out of the engagement with a platform that was faster, more secure, and meaningfully cheaper to run. The 80% reduction in cloud infrastructure cost is the most concrete number on file, and it holds up as a cloud cost optimization case study precisely because the savings were paired with 130+ shipped user stories rather than achieved by cutting scope. In effect, the GCP audit paid for the rest of the project — the cost savings gave the engagement room to deliver full product depth without trading one priority for the other.
Christopher Massood, co-founder and CEO of ElephantCPA, has been on record on Clutch describing the engineering work as transparent and honest, with the team going to extra lengths to maintain a clean process.
"Leanware is transparent, and they’re very honest with the work that they do."
— Christopher Massood , Co-Founder & CEO , Elephant CPA · Wayne, New Jersey
Engagement FAQ
What are cloud cost optimization services and how do they work?
Cloud cost optimization services review an existing cloud setup (compute, storage, and how resources are provisioned) to find where a company is paying for capacity it does not use. For ElephantCPA, a Cloud Advisory review of the GCP infrastructure found significant over-provisioning, and applying the recommendations across compute and storage cut cloud spend by 80%.
Can you reduce cloud costs without hurting platform performance or security?
Yes, when the review is tied to the actual architecture rather than just the billing dashboard. ElephantCPA's cost reduction came out of the same architecture overhaul that tightened the platform's cybersecurity posture and stabilized the system end to end, so the 80% savings did not trade off against reliability.
What does GCP cost optimization look like for a SaaS platform handling sensitive data?
It starts with a review of how compute and storage are actually being used versus how they're provisioned. ElephantCPA's GCP cost optimization work came out of a Cloud Advisory review that found significant over-provisioning across the platform's infrastructure, and the recommendations, applied without narrowing the product roadmap, cut cloud spend by 80%.
What does accounting software development typically involve for a SaaS platform handling financial data?
It covers the full surface an accountant-client platform needs: account management, document handling, secure messaging, scheduling, billing integrations, and reporting. ElephantCPA's engagement shipped 130+ user stories across those areas, integrated with Stripe for billing, SendGrid for email, and Zoho for CRM.
How is a cloud cost audit different from a general infrastructure review?
A cloud cost audit focuses specifically on provisioning: which compute and storage resources are sized for load that does not exist. ElephantCPA's Cloud Advisory review was scoped this way and is documented as a cloud cost optimization case study because the 80% reduction is a verified, on-record outcome rather than an estimate.
What team size is needed to enhance an existing SaaS platform while also cutting cloud costs?
ElephantCPA's enhancement ran with a senior full-stack engineer leading, a mid full-stack engineer on feature work, a product designer, and a product owner, four people total. The cloud cost work was folded into the same architecture overhaul rather than staffed as a separate workstream.