How iManage reinvented their SaaS future with AxonIQ

When iManage set out to rebuild one of their core products for the cloud, they weren’t just migrating systems — they were chasing something bigger: agility, speed, and trust at a global scale.

Trusted by many of the world’s top law firms and financial institutions, iManage powers secure document and email management for industries where information governance isn’t just important — it’s everything. In the world of SaaS, where every second matters and every mistake compounds, clinging to legacy architectures isn’t just inefficient — it’s dangerous.

They needed a foundation that would scale with them — one architected for today’s cloud-native expectations and tomorrow’s unknown demands.

When the stakes were highest, they chose Axon.

For years, iManage maintained a powerful but increasingly tangled on-premises system. Originally born from a startup acquisition, the platform’s extreme configurability eventually became its Achilles' heel.

"People started building features we didn’t really want them to be building," Tom explained. "If there was ever a bug, engineering would be reverse engineering the configurations, trying to find out what they were trying to achieve."

Each customer deployment became its own unique snowflake — slightly different, slightly incompatible, slightly unstable. Support tickets ballooned. Performance tuning became guesswork. Rolling out improvements across customers was almost impossible. Meanwhile, the system relied heavily on synchronous processing, which is a bottleneck when dealing with massive document indexes."Processing a document and indexing it into a large search engine is quite slow," Tom said. It doesn't suit a synchronous architecture."

Security risks loomed larger in a SaaS world. The old system let customers configure database and file system connectors themselves — acceptable on-premises, but a serious exposure risk in the cloud. Maintenance was another daily grind. Legacy Java frameworks like OSGi made even basic upgrades painfully slow. In some cases, parts of the platform were effectively frozen in time.

Scaling the monolith to meet cloud-native SaaS demands wasn’t just impractical — it was fundamentally unsafe.

The choice became clear: patch endlessly, or rebuild fearlessly.

iManage chose the latter.

With a clean slate, Tom led a months-long research effort into building a system that could truly meet SaaS demands: high throughput, resilience, elasticity, and operational simplicity. An asynchronous, event-driven architecture became the obvious path.

Document indexing and metadata extraction were slow, distributed processes — they needed services that communicated reliably but loosely. Reliability was critical. Missing a security event could have disastrous consequences. "We needed to guarantee delivery," Tom emphasized.

Tom tested multiple stacks:

• Axon Framework + SQL Server for event storage.
• Axon Framework + Kafka for messaging.
• Other event sourcing frameworks he encountered through books and open-source projects.

Each option brought complexity: more bootstrapping, manual queue management, fragile integrations. "You have to figure out what queues you need, what events go on which topics, how you tie publishers and subscribers together. It becomes a lot to manage," Tom said.

The turning point came when he paired Axon Framework with Axon Server.

Axon Server handled persistence, messaging, and recoverability seamlessly — without forcing developers to build and maintain custom broker architectures.

The decision was clear. iManage chose Axon.

The technical shift was massive. But the cultural shift was even bigger. Instead of huge, tightly coupled deployments, services were broken down by aggregate ownership. Instead of big bang releases, iManage moved to daily CI/CD pipelines, feature flags, and a clear separation of deployment vs release.

"We stopped thinking in terms of massive releases," Tom said. "Now, deployment and release are two different things." And all of this transformation happened remotely. The COVID-19 pandemic meant engineers weren’t gathering in rooms — they were distributed across London, Bangalore, and Chicago, collaborating entirely through screens. "Even with remote work, event sourcing clicked faster than expected once people got into it," Tom noted. AxonIQ training grounded teams in concepts first — aggregates, commands, projections — and tooling second. Most engineers only needed to learn how to write clean, annotated handlers. A handful became Axon experts who handled deeper troubleshooting and architecture.

The architecture itself — decentralized, event-driven, scalable — mirrored the global, decentralized team building it.

As iManage’s new platform grew, so did operational complexity. Today, iManage runs:

• 10 independent Axon Server clusters across development, staging, pre-production, and production environments.
• Hundreds of Axon Framework-based microservice instances are deployed separately for each tenant.
• 30+ production tenants, each with 60–90 services running simultaneously.

Monitoring this spider web was nearly impossible without centralized tooling. That's where AxonIQ Console became indispensable. "Having AxonIQ Console where everything is pushed into one application — even though it's separated per cluster — is really helpful," Tom said. "We're able to have sensible alerts set up by people who actually know what should be measured and what looks iffy."

Before AxonIQ Console, the team pushed service metrics into Grafana dashboards — but with separate Grafana instances per region and level, global visibility was fragmented. AxonIQ Console unified their view across all clusters. More importantly, AxonIQ Console gave iManage operational control:

• Resetting event processors across distributed services without needing manual Kubernetes hacks.
• Splitting and merging segments easily to rebalance loads.
• Pausing and resuming processors in synchronized, cluster-wide operations — something extremely difficult under traditional HTTP load balancing.

AxonIQ Console also strengthened data security:

Unlike raw Axon Server UI access, AxonIQ Console allowed engineers to manage system health without exposing sensitive event data — supporting GDPR compliance and internal role-based access control policies. Today, iManage is the largest AxonIQ Console user worldwide — and it’s a critical piece of their operational resilience.

Today, iManage’s cloud-native SaaS platform operates with the reliability and agility their customers demand.

• Support incidents have dropped dramatically.
• Engineering teams move faster and more independently.
• Onboarding new tenants is faster, safer, and less risky.
• Operational visibility and control have never been stronger.

Concerns about event store growth have proven manageable with real-world data and pruning strategies. Most importantly: engineering can focus on building legal tech solutions — not managing messaging systems, databases, or monitoring infrastructure.

"We’re not in the business of managing databases," Tom said. "We’re in the business of building legal software. Axon lets us focus on what we’re good at."

iManage’s journey offers clear lessons for SaaS builders:

• Start clean when you can. Avoid dragging old-world tech debt into the cloud.
• Build for resilience and recovery. Event-driven systems naturally scale and self-heal.
• Use purpose-built tools. Focus your energy where it matters
• Centralize observability early. Unified monitoring and operational control prevents firefighting later.
• Prioritize focus.Let your team innovate on your core product — not on plumbing, brokers, or homegrown monitoring hacks.