From our first experience using Cleric, it was totally different and much better than anything else we’d been testing. It was obvious that this was built by platform engineers.
Amina Odžak Engineering Manager at Mixmax
How Mixmax reduced incident resolution time by 90% with Cleric
84% Reduction in alert noise
60+ Tickets resolved in one week
90% Reduction in incident resolution time 39 hrs (before) → under 4 hrs (after)
On Mixmax’s engineering team, triage and investigation used to happen manually. Now, when there’s a support ticket or on-call alert, Cleric immediately digs in and provides a diagnosis of the problem. Then an engineer resolves the issue with Claude or other tooling, after verifying Cleric’s analysis.
“Honestly, though, people don’t double-check Cleric’s analysis as much anymore,” says Amina Odžak, Engineering Manager. “Cleric is really good, and the team trusts it.”
Mixmax helps teams of all sizes with outreach, scheduling, and sequencing. AI has changed their industry quickly, and their engineering team has adapted too, finding new ways to deliver improvements to customers sooner. “With the agentic way of working, we’ve doubled, if not tripled our output,” Amina says. “We’re delivering much faster than ever in the history of the company.”
With Cleric, Mixmax can stay focused on development rather than noisy alerts, and fix customer issues in minutes (often without even involving the engineering team).
More output with agents, and more operational noise
As Mixmax increased their PRs 2–3x, they saw a corresponding increase in alerts and production issues. The noise was becoming impossible to manage, and the volume of code had outpaced any one person’s ability to understand the system. “We’ve standardized around internal and external MCPs, but each developer has their own way of working—they might do something in an entirely different way than someone else on the team,” Amina says.
When they were on-call, engineers often encountered parts of the product they’d never touched. They might struggle to understand how to solve a problem or export information for a customer. And because of timezone differences, it could take more than 24 hours for them to get the necessary context from coworkers and resolve the ticket.
With so much fragmentation and speed on the development side, what the team needed was better production observability. They wanted a tool that fit their way of working, that could reliably suppress noisy alerts while providing a fast resolution path for real issues.
Why Mixmax chose Cleric
Amina and her team researched multiple third-party tools for production observability and alert investigation. It came down to Cleric and one competitor.
“We thought we could extend usage of another tool in our stack,” Amina explains. “But it’s very noticeable when something is just a feature, versus when someone builds a whole product like Cleric that focuses on doing one thing really well.”
They saw a difference in quality as well as cost: “The other provider was also more expensive compared to what we needed from it in the end,” Amina says.
Cleric proved its value when the team had a production incident involving paid accounts being mislabeled as free-tier. This was about as visible as a bug gets, since it directly affected whether customers could use features they were paying for.
Cleric caught it and filed the ticket itself within a minute of the incident being reported. Over the next two hours, Cleric revisited the incident across several investigations, refining its diagnosis as new evidence emerged. Ultimately it found that the root cause was underprovisioned compute causing entitlement checks to time out.
“One of our engineers validated the fix and shipped it, and paying customers were back to full access within an hour. What stood out to us is that Cleric did the legwork of ruling things in and out, while a human still made the final call, which is really the workflow we wanted,” Amina says.
Statuspage 13:57 Entitlement checking issues
Some users are experiencing issues getting access to the features they are subscribed for -- their accounts behaving as free user accounts.
Completed 4 investigation steps
16:17 What I found:
- Root cause (confirmed): users-api production ECS tasks were CPU-under-provisioned → event-loop starvation → ALB health-check timeouts → ~2h crash-loop stalling the lever/entitlement endpoint. billing-features’ getUserLever(usageBillingM1) calls timed out and resolved paid accounts as free.
- Timeline: outage 12:35Z → 14:55Z. ELB 5xx peaked at 1.34M (13:20Z); p99 hit 25.5s.
Overall, as the team piloted Cleric they were impressed by its simplicity, ease of use, and accuracy compared to competitors. Mixmax Platform lead Ahmed Anas says: “Cleric’s AI quality is in line with any state of the art AI models, but where it excels is providing an unrivaled user experience. In any alert, or any part of our chat, ticket, etc., it automatically runs or easily allows us to start an initial investigation, finds the cause and severity of the situation, and reduces the mental strain to start a new line of work. The integrations are easy to set up, and it has many, so it can easily pick up the necessary information to provide an informed and accurate analysis.”
Results
With Cleric, alert noise has been reduced 84% (from 1180 alerts to 190 action-required issues that Cleric surfaced over the past 5 months).
The team has also seen a huge change in the MTTR for incidents. “In the past where we needed a couple of days, now we’re often resolving tickets in a matter of minutes,” Amina says. The average incident time has dropped from 39 hours as measured in the quarter before adopting Cleric, to under 4 hours in the quarter after adopting Cleric—a 90% reduction in incident resolution time. “That’s even in a quarter where more incidents were coming through,” Amina says.
During a recent sprint to clear its customer-support backlog, Mixmax resolved more than 60 tickets in one week. “I know the team was relying on Cleric to find the root cause and come up with resolutions,” Amina says.
Fewer alerts reach an engineer, and the ones that do are resolved in hours
84%
of alerts were triaged as noise
190 needed an engineer 1,180 fired
Over five months
90%
less time to resolve an incident
Under 4 h with Cleric 39 h before
Average per incident, a quarter before and after
The customer support team itself uses Cleric on a daily basis. Previously, they often needed to rope in engineering to understand the technical aspect of a customer issue. Now, they’re frequently able to take Cleric’s explanation of the problem and communicate it back to the customer. “We’ve reduced the number of customer service tickets that come to engineering because that team now has the tooling to move faster and analyze issues with confidence,” Amina says.
A trusted partner
Beyond the quantitative impact, one of the things that stands out for Amina is the experience of working with the Cleric team. “The co-founder, Shahram, would show up to our onboarding meetings just to see how we were doing. He’s down-to-earth and genuinely cares, and somehow everyone around him has the same personality. You can tell they want to provide great value and improve the product any way possible.”
Mixmax now wants to extend Cleric’s role from investigation into automated resolution. As Amina puts it, “We’d like to get to the point where Cleric does the analysis and our Claude agent picks it up and implements the fix. We’re not quite there yet, but we’re almost there with Cleric and its capabilities.”
Give your on-call a head start.
Start for free, or talk to us about a plan built for your team’s scale and security needs.