How Ada used Doppler to boost efficiency for secrets management by 95%

Ada is an AI-native customer service platform built for enterprises to deploy AI agents with transparency, control, and confidence, delivering customer service that is accurate, safe, and on brand across channels. They use large language models to resolve inquiries instantly, scaling exceptional customer experiences without increasing headcount.

As Ada grew and focused more on enterprise-readiness, the team wanted a more efficient method to manage secrets. They adopted Doppler to reduce friction, speed up development, and gain built-in workflows for securely managing production secrets.

Growth outpaced internal secrets management

Ada started with a homegrown secrets tool utilizing KMS encryption and a Git-based system built for a small, colocated team. When everyone worked in the same office, it was enough.

As the team grew and shifted to remote-first development, they needed an agile solution with more efficient processes.

Workarounds became common, and secrets management ended up decentralized and inconsistent across multiple tools and teams. Infrastructure engineers struggled to enforce consistency and maintain visibility.

We are not a secrets management company, we’re an AI customer service company. We wanted to spend less time managing secrets and more time on our core dev work.

Vaults fell short. Doppler didn’t.

Ada evaluated Vault, AWS Secrets Manager, and Akeyless. Each offered the basics of secret storage, but none delivered a better developer experience, which was a high priority. Most options still required the team to build workflows, create internal UIs, and maintain glue code.

Doppler provided all of that out of the box. The interface, CLI, access controls, and approval flows were already built and ready to use.

“Most of the other tools, like Vault, AWS Secrets Manager, and Akeyless, were just vaults. There was no solution that just made it easy for people to go in, change secrets, and edit them to get what they need and then get out of there.”

After hearing about Doppler on a Changelog podcast, they ran a successful proof of concept before rolling it out to where it could have the most impact.

The team led with its most critical service

Rather than start small, they integrated Doppler into their largest and most complex Python monolith, which had been part of their architecture since the company’s beginning. This application had the highest risk and the highest value.

“We didn’t roll it out slowly. We started with the largest, hardest service because that’s where it mattered most.”

That success gave the team a repeatable pattern. Every new service now includes a Doppler project by default, with access provisioned through Okta.

Change Requests brought security and speed together

Before Doppler, production secret updates relied on GitHub pull requests, Slack messages, or approval tickets. The process was slow and fragmented, taking hours and 19 steps to change a single secret.

Doppler’s Change Requests feature solved the problem directly.

“Secrets are code. You need change management around them. Doppler gave us that without building our own tool.”

Now, developers request changes in Doppler. Team leads review and approve them inside the same system, ensuring every update is tracked and auditable.

Secret updates no longer slow anyone down

After adoption, the team set up a Slack integration to monitor Doppler activity. Usage increased right away.

“We saw that people were doing more secret changes than before. And they were doing it securely. We increased usage. We increased how many times people did the process because it wasn't a bottleneck anymore.”

Ada now manages all secrets through Doppler, including those used in CI/CD pipelines and infrastructure provisioning via Terraform. Engineers no longer rely on institutional knowledge, migration has become self-serve, and the support burden has dropped.

Doppler is now the standard

Today, all of Engineering relies on Doppler. It’s fully integrated into Ada’s internal developer platform, automatically provisioned for every new service, and used daily by developers, team leads, and infrastructure engineers.

“Doppler keeps raising the bar on secrets management. It’s obvious in hindsight.”

To reduce human-managed secrets, the team is expanding its use of IAM roles, machine-to-machine auth, and OIDC. Doppler’s user-based pricing supports this shift by allowing unlimited machine identities at no additional cost.

“We now have one way to manage application configuration that anyone can learn in less than an hour after joining our company.”

Next, Ada plans to expand their use of Doppler Share, Secrets Rotation, and Dynamic Secrets to further reduce manual work and improve security.