CONTACT US
CONTACT US

Reduced MTTR by 20% with AI-Powered Diagnostics for a Global Automotive Company

At a glance

  • Client
Global Automotive Manufacturer
  • Industry
Auto & Manufacturing
  • Solution Provided
AI-powered diagnostics, automated root cause analysis, agentic workflows for SRE support, observability integration
  • Technologies Used
Construct™: Diagnose, Kubernetes (GKE), Apache Airflow, Google Gemini, Agno Framework, Prometheus, Google Cloud Platform (GCP), ServiceNow
The Challenge

A global automotive manufacturer experienced a critical infrastructure outage that brought key services offline, preventing customers from unlocking their vehicles via mobile and halting credit application processing. Operating a massive orchestration platform with hundreds of Airflow DAGs on Kubernetes, the company faced rising challenges around operational reliability and incident response.
Internally, manual triage across distributed systems made it difficult for platform engineers to isolate issues quickly, respond to incidents efficiently, and maintain focus on platform innovation. Recurring failures slowed resolution times and strained engineering capacity.

Key challenges included:

  • High-severity outages with wide impact: Failures affected consumer-facing features and core backend processing, prompting executive-level escalation and overnight war rooms with 400+ staff.
  • Manual failure investigation: Engineers had to cross-reference logs and metrics across Airflow, Kubernetes, and system components, slowing root cause analysis and increasing MTTR.
  • Growing support load: Recurring issues flooded platform support queues, pulling engineers away from roadmap work and reducing productivity.
  • Lack of diagnostic automation: No centralized system existed to correlate observability signals or assist engineers in real-time troubleshooting.
The Solution

To improve operational resilience, Gorilla Logic deployed an AI-enabled diagnostic solution that combined observability integration, agentic reasoning, and workflow orchestration. This solution wasn’t a one-off intervention—it was enabled by Construct™: Diagnose, one of several delivery-tested workflows that make up *Gorilla Logic Construct™.

Key elements of the solution included:

  • AI-Driven Diagnostic Agent: Built on the Agno framework and powered by Google Gemini, the agent dynamically interprets queries and performs multi-step reasoning across observability data.
  • End-to-End Signal Correlation: The agent automatically correlates logs from Airflow, events from Kubernetes, metrics from Prometheus, and traces from GCP to uncover the root cause of failures.
  • Workflow Integrations: Seamless connections to GKE, GCP, Airflow, and ServiceNow enable incident resolution to be fully embedded into the client’s operational processes.
  • Natural Language Interface: Engineers receive real-time, conversational diagnostics via natural language prompts, simplifying usage and accelerating response.

The Results

20% Faster MTTR: Incidents resolved more quickly, reducing downtime and restoring customer functionality faster.
Up to 30% Engineering Ablation: The agent handled 20–30% of routine diagnostics, freeing engineers to focus on platform innovation.
Improved System Reliability: Accelerated identification of root causes across pipelines and infrastructure.
Scalable AI Model: A reusable, agentic diagnostic architecture now available for broader rollout.

*Gorilla Logic Construct™ is how we deliver faster—with less engineering lift and greater confidence.

It’s not a product. It’s our portfolio of delivery-tested workflows, powered by modular AI agents. Every workflow is proven in delivery, reusable by design, and capable of cutting engineering work by 30-80%.
Construct™: Diagnose is our issue intelligence agent—built to diagnose root causes in real time to accelerate resolution.