Use Case Intelligence RSS Search use cases Search
Google Cloud Software / SaaS India

InMobi

Application Refactoring · Real-time Analytics · System Migration

InMobi needed to migrate a latency-sensitive advertising platform from another hyperscaler, preserve zero-downtime service at massive QPS, and create a more efficient real-time serving and analytics environment for both auction traffic and future GenAI workloads.

Read the original story Explore all use cases
7 million peak qps
Lowered overall infrastructure costs by 15%
Reduced recovery time to ~10 mins
performing >50M ML inferences daily with zero downtime

Reported impact

Business outcome

InMobi completed a live migration without downtime, reduced cost, and created a more elastic foundation for ad-serving analytics and future GenAI advertising features.

  • Cost Reduction
  • Scalability
  • Speed to Market

Solution

InMobi executed a staged migration to Google Cloud that combined core-serving modernization with a large analytical data plane. Google Kubernetes Engine now hosts containerized real-time serving workloads, using custom ComputeClasses and autoscaling features to react to volatility in spot capacity. Cloud Storage holds petabytes of supply-side platform data, Hyperdisk provides low-latency block storage for performance-critical paths, and Databricks on Google Cloud plus Cloud Composer support distributed processing and hundreds of orchestrated data pipelines. The company also mapped additional workloads onto Compute Engine and Cloud SQL during the migration, then used Bigtable and in-memory caching in the serving layer to keep ad-delivery latency low. The result is a platform that can process millions of queries per second and tens of millions of ML inferences daily with zero downtime while lowering infrastructure cost and setting up future agentic advertising work on Vertex AI.

Data flow

Ad-serving requests hit GKE-hosted services in the real-time path, where cached and low-latency data layers help return responses quickly. Large historical datasets live in Cloud Storage and flow through Databricks and Cloud Composer pipelines for analytics, model development, and future Vertex AI workloads.

Solution architecture

6 components · 4 layers
  1. Compute
    • Databricks on Google Cloud Executes distributed data-processing workloads on the migrated platform.
  2. Storage
    • Cloud Storage Stores petabytes of supply-side platform data for analytics and model-development workflows.
    • Google Cloud Hyperdisk Provides consistent low-latency block storage for performance-critical serving and data-pipeline workloads.
  3. Serving
    • Bigtable Supports the serving flow with fast-access data needed by real-time ad delivery.
  4. Orchestration
    • Google Kubernetes Engine (GKE) Runs InMobi's real-time serving workloads and dynamically scales capacity under live auction traffic.
    • Cloud Composer Orchestrates hundreds of production data pipelines in a managed Airflow environment.

Architecture clues

  • Cloud Composer orchestrates hundreds of production data pipelines
  • GKE custom ComputeClasses manage capacity across instance types
  • InMobi plans to extend the stack into full agentic advertising on Vertex AI
  • Production traffic was migrated incrementally during the cutover

Evidence from the source

As the data layer, Cloud Storage houses several petabytes of supply-side platform data
At the heart of the serving architecture, containerized workloads run directly in the real-time serving path
Lowered overall infrastructure costs by 15% on Compute Engine versus another hyperscaler
Processed 7 million peak qps performing >50M ML inferences daily with zero downtime
Spark workloads run on GKE, Databricks operates on Google Cloud for distributed processing, and Cloud Composer orchestrates hundreds of production data pipelines
Transparency Evidence & methodology 95/100 reference value · 79% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value95/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth20/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact25/25
    2 metrics describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence79%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability21/25
    10 of 12 key fields link to a named source.
  • Direct evidence20/20
    5 direct source quotes retained.
  • Date provenance0/10
    The publication date has no documented provenance.
  • Source transparency8/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. Google Cloud

    InMobi customer story

    First-party
    https://cloud.google.com/customers/inmobi
    Source type
    Vendor-published customer story
    Supports
    • Customer
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes

Date evidence

  • First archived Jun 11, 2026 · Internet Archive · Independent First Wayback Machine capture; confirms the page existed by this date, not when it was published.
    Open date source
  • Sitemap last modified May 19, 2026 · Google Cloud · Publisher metadata Publisher-reported modification timestamp; not a publication date.
    Open date source
Not separately sourced Industry, Region, Company size. Treat these normalized attributes as lower-confidence until a dedicated source is added.

Related

Similar use cases

See all like this
Google Cloud Est. · Apr 13, 2026 Shopify Retail North America Shopify powers millions of merchants in more than 175 countries whose traffic can spike within seconds — particularly during Black Friday and Cyber Monday — and traditional static environments would have required costly… $14.6B in global Black Friday-Cyber Monday weekend sales (27% YoY) Application RefactoringIT Operations Automation Compute Engine · Google Kubernetes Engine · Hyperdisk Balanced Google Cloud Archived · Aug 29, 2026 WeSpeak Software / SaaS Argentina Building that product required infrastructure that would never go down. Early on, WeSpeak ran on a virtual private server that could fail without warning, taking the entire platform offline. For a hotel waiting on a… 70% of hotel guest inquiries automated end to end Application RefactoringIT Operations Automation Cloud Run · Gemini · Gemini Flash Google Cloud Archived · Apr 13, 2026 Tictag.io Software / SaaS Singapore A private endpoint within Virtual Private Cloud (VPC) that connects the company's cloud environment securely and with low latency to its primary MongoDB Atlas database, and To power intensive AI processing for its… "We selected the NVIDIA L4 Cloud GPU because we can fit two into one node and it sits at the sweet spot of cost… Application RefactoringReal-time Analytics Cloud CDN · Cloud GPUs · Cloud Storage