How Labra Moved Beyond New Relic to Get a Simpler, AI-Powered Observability Experience with Oodle

“We just say — Hey, this is my service, this is my request ID, get me the RCA. Boom, you have it, along with the source artifacts it referenced.

Oodle AI Assistant saves a lot of time for us.”
Priyank Raj
Senior Software Engineer, Labra
1.5 hrs
50-Panel Dashboard Built
AI-First
Debugging & RCA
3 Clouds
Unified Observability
<2 Weeks
Onboarding + Migration
Labra Logo

Labra powers cloud go-to-market for SaaS companies, helping them list, sell, and manage software across AWS, Azure, and GCP marketplaces.

IndustryCloud Commerce / SaaS
Company size51-200
Previous ToolsNew Relic, CloudWatch

Challenge

“As we scaled across three cloud marketplaces, our observability needs grew fast but our tooling couldn't keep up. We needed something that could match our pace without requiring a dedicated platform team to maintain it.”
Nestor, Principal Engineer, Labra
“New Relic's per-user pricing model made it difficult to give every developer the access they needed as the team grew.”
Priyank Raj, Senior Software Engineer, Labra

Labra's observability stack had evolved alongside rapid product growth. CloudWatch handled AWS logs, but as the team expanded across Azure and GCP, they needed APM and distributed tracing. New Relic was added to fill that gap but rather than simplifying operations, it introduced new complexity and cost without delivering the multi-cloud visibility the team needed.

Pain Points:

Complicated alerting: Alert configuration was so complex that setting up granular alerts was extremely difficult.
Per-seat pricing at odds with team scale: Prohibitively expensive for every developer to have the access they need.
Redundant costs: New Relic's reliability gaps made running CloudWatch in parallel unavoidable.
Limited ROI on New Relic: Its complex interface made advanced features effectively out of reach without significant ramp-up time.
5-minute Azure log delay: Real-time debugging was impractical for a multi-cloud team.
Proprietary vendor lock-in: Instrumentation tied to New Relic's SDK. Any vendor change meant re-instrumenting everything.

Solution

Why Labra Chose Oodle

“We were really impressed by the AI debugging. We built a custom agent on the Oodle MCP and the AI felt more mature in Oodle than anything else we evaluated.”
“Oodle is an application with almost every feature at a cheaper price. It is also simpler and easier to debug.”
Priyank Raj, Senior Software Engineer, Labra

The PoC started small: three QA ECS services connected via an OpenTelemetry collector running as a sidecar container. The onboarding was seamless.

What made the difference:

AI-native debugging: AI assistant and MCP server for PromQL, alerts, dashboards, and automated RCA.
Simpler, intuitive UI: Features previously unused became accessible with a cleaner interface.
No per-user fees: Pricing scales with data, not headcount. Every developer gets access.
OTel-native integration: No proprietary SDK lock-in. Standard OpenTelemetry across all services.
Responsive support: Slack replies in minutes. Features shipped during the PoC. Oodle University enabled self-serve onboarding.

Impact

From PoC to Production

“It took hardly 1.5 hours of prompting to create a 50-panel dashboard. Manually, it would take almost 1.5 days. But Oodle did it in one stretch.”
“Oodle is pretty much AI-first. Our devs opened the logs page to debug production and found the issue and RCA right in the AI Summary — without having to sift through a single log.”
Priyank Raj, Senior Software Engineer, Labra

Within a week, developers were using Oodle for real debugging on migrated services: searching logs by request ID, tracing requests across services, and building dashboards for marketplace transaction metrics.

50-panel dashboard in 1.5 hours: AI assistant replaced a day and a half of manual dashboard building.
One platform across three clouds: Logs, metrics, and traces from AWS, Azure, and GCP in one place.
Automated RCA on every alert: Custom LangChain agent integrated with Oodle MCP to automate alert investigations.
Trace heatmap for latency outliers: Spot latency anomalies at a glance, no queries needed.
Synthetic monitoring at a fraction of the cost: Replaced CloudWatch Synthetics, delivered days after the request.
“When alerts are triggered, our custom agent connects to the Oodle MCP, runs the investigation, and gets the RCA with artifacts.”
Priyank Raj, Senior Software Engineer, Labra

Experience observability that lets you focus on what makes your business unique.

Request a Demo →