Constellation Gate AI | Customer case study
One of the web's highest rated resume builders routes its AI requests through Constellation Gate AI, using built-in caching and compression to reduce token spend and an immutable audit log to see exactly how every AI token is used across software development.
Rocket Resume is a high-traffic, AI-assisted resume builder that generates and refines resumes for hundreds of thousands of job seekers every month. By routing its AI message and requests workload through Constellation Gate AI, Rocket Resume reduced the cost of that spend through built-in caching and prompt compression, gained a tamper-evident record of how every AI token is consumed, and turned that visibility into an internal program that rewards engineers for effective AI use.
Rocket Resume (rocket-resume.com) is one of the highest-rated online resume builders, helping job seekers create, edit, and tailor professional resumes in minutes, with AI generating and refining content throughout the process. The platform attracts approximately 150,000 visits per month, with traffic growing nearly 28 percent month over month. Visitors average more than 14 pages per session, reflecting the highly interactive nature of the resume-building experience. With a major redesign, further international expansion, and new product launches on the horizon, Rocket Resume is entering its next phase of growth.
A product where nearly every session invokes a large language model has two problems that grow in lockstep with success.
As traffic climbed, so did development and token consumption, reaching roughly $40,000 per month. Much of it was avoidable: repeated similar requests, prompts heavier than the model needed, and no cost-reduction layer between the application and the providers.
GitHub shows commits, not how AI is consumed. The team needed an independent, trustworthy record of which engineer and which product was using tokens, and confirmation that spend mapped to sanctioned work.
Rocket Resume adopted Constellation Gate AI, a self-serve LLM gateway that sits inline between an application and over 300 available models. Two capabilities were central to the deployment.
Gate AI's inline cost-reduction and compression layer cuts upstream token spend without any changes to application code. Repeat-request caching matches an incoming request against the ones it has already answered, so an identical call is served from the stored response without paying for a fresh model call. Prompt compression sends repeated content once and points at it afterwards, reducing the tokens billed on the way in. Savings are reported per workspace, so Rocket Resume can verify the reduction directly against its provider bills.
Every request that passes through Gate AI is written to a tamper-evident audit trail anchored to Constellation's Digital Evidence layer. This record cannot be quietly edited after the fact; it is cryptographically anchored, so what happened is independently verifiable. Rocket Resume uses it as an engineering observability system that goes beyond GitHub: seeing how tokens are actually consumed, attributing usage to the right people and products, and confirming tokens are not leaking across unrelated engineering projects.
“As we add AI-powered tools, we need to understand both what those systems contribute and what they cost,” said Steve Zimmerman, founder of Rocket Resume. “Gate AI gives our engineering team a clearer view of the workflows we’ve connected and helps us identify opportunities to operate more efficiently without compromising the experience we provide to job seekers all while saving money!”
“This is exactly why we built Gate AI. Any company running real AI workloads faces the same two questions: how do I control what this costs, and how do I know what my systems and my team are actually doing with it. Rocket Resume answered both without changing a line of their application, and then took it a step further by rewarding their engineers for using AI well. I am proud of what our team built here, because it turns AI from an unpredictable expense into something a company can measure, trust, and improve.”
Rocket Resume is a preview of how software companies will run on AI: with a cost layer that keeps token spend in check and an audit layer that makes AI usage transparent and verifiable. Constellation Gate AI delivers both in a single drop-in gateway, with the same cryptographic integrity that anchors the rest of the Constellation ecosystem.
A self-serve LLM gateway that sits inline between an AI application or agent and the model providers. It adds tiered prompt-injection defense, output-side scanning for PII and credentials, inline token cost optimization through caching and compression, and a tamper-evident audit trail anchored to Constellation's Digital Evidence layer. Available as a free tier, a paid subscription, and a pay-as-you-go managed-key option, all self-serve with no procurement loop.
Start on the free tier, point your application or your agent at Gate, and every request after that is screened, compressed, and recorded.