Case Studies › ExpressVPN

Cybersecurity & Privacy SaaS

Scaling Privacy Infrastructure for Millions Globally

Techdots re-engineered ExpressVPN's subscription and device orchestration layer to withstand global traffic spikes without service degradation. We built a distributed session management system, a real-time diagnostics pipeline, and a self-service account portal that dramatically reduced churn and support load. The result is a platform that quietly serves millions of privacy-conscious users across 160+ server locations with near-zero downtime.

Start a Similar Project → ← All Case Studies

ExpressVPN

Cybersecurity & Privacy SaaS

99.97%

API uptime post-launch (up from 99.1%)

4.2x

Faster connection handshake under peak load

38%

Reduction in support tickets via self-service portal

20 weeks

timeline

This engagement is best for

Consumer SaaS platforms with millions of concurrent active sessions

Security or privacy products where downtime directly erodes user trust

Subscription businesses needing robust device and entitlement management

Companies with global infrastructure struggling to scale support operations

The Transformation

Before & After

Before

✕ Subscription service APIs degraded noticeably during promotional traffic spikes, causing failed logins and connection drops

✕ Device entitlement logic was embedded in a monolithic Rails app with no separation of concerns, making changes risky

✕ Customer support was flooded with password reset, device limit, and billing confusion tickets — no self-service path existed

✕ Real-time diagnostics were unavailable; engineering relied on lagging CloudWatch metrics to detect regional outages

✕ Renewal and cancellation flows were fragmented across three codebases, causing inconsistent user experiences and revenue leakage

After

✓ Subscription APIs sustained 3M+ concurrent sessions during a major campaign launch with zero degradation

✓ Device and entitlement logic extracted into a dedicated microservice with a clean gRPC interface consumed by all client apps

✓ Self-service portal reduced inbound support volume by 38% within 60 days of launch

✓ Real-time telemetry pipeline surfaces regional connection anomalies within 90 seconds, enabling proactive incident response

✓ Unified renewal and cancellation flows increased annual plan conversion by 11% and reduced involuntary churn

What We Built

Deliverables & Scope

Every item below was chosen because it directly addressed a business bottleneck — not because it was technically interesting.

Distributed session and entitlement microservice in Go, exposing a gRPC API consumed by iOS, Android, Windows, and Mac clients

Event-driven telemetry pipeline using Kafka and ClickHouse to aggregate connection health signals across 160+ server locations in near-real-time

React-based self-service customer portal covering device management, billing history, plan changes, and cancellation flows

Automated dunning and renewal orchestration system integrated with Stripe Billing, handling retry logic and prorated upgrades

Load-tested chaos engineering suite using Gremlin to validate failure modes across simulated regional outages

CI/CD pipeline with blue-green deployment support on AWS EKS, enabling zero-downtime releases across all services

ROI Logic

Why This Generated
Real Business Value

Every minute of degraded service for a privacy product translates directly into churn — users who cannot connect during a sensitive browsing session switch providers and rarely return. By hardening the subscription and session layer, Techdots eliminated the failure modes that had been quietly costing ExpressVPN renewals. The self-service portal alone offloaded enough support volume to recoup the engagement cost within two billing cycles.

Key Outcomes

99.97%

API uptime post-launch (up from 99.1%)

4.2x

Faster connection handshake under peak load

38%

Reduction in support tickets via self-service portal

Why It Worked

The Decisions That
Made the Difference

Good execution matters. But the right early decisions matter more.

We extracted entitlement logic before touching the API layer, giving every client surface a single source of truth and eliminating the class of bugs that came from duplicated business rules

The telemetry pipeline was designed to be append-only and schema-flexible from day one, so new signal types could be added without migrations or downtime

We ran the self-service portal in shadow mode alongside the legacy flow for three weeks, validating parity before cutting over — preventing any surprise gaps in edge-case billing scenarios

Chaos engineering was integrated into the staging environment as a required gate before every production deploy, not treated as a one-time audit

Tech Stack

Apache Kafka ClickHouse AWS EKS

Integrations

Stripe Billing (subscription lifecycle and dunning) Gremlin (chaos engineering and fault injection) PagerDuty (real-time incident alerting from telemetry pipeline) Braze (lifecycle messaging for renewal and cancellation flows)

Start your project

Have a Similar Problem?

Start with a Software + AI Audit. We'll map your workflows, identify the highest-ROI opportunities, and give you a clear roadmap before you commit to development.

Get a Software + AI Audit → ← View All Case Studies