Challenge
Insight alone does not change outcomes — decisions and actions do. Human Managed’s customers in finance, government, and education needed guidance that could keep pace with more than 40 telemetry sources feeding one framework.
Early AWS usage was infrastructure-as-footing. As volumes and expectations rose, the limiting factors became latency to decision, orchestration across services, and operational load on Human Managed’s own platform team — not willingness to automate.
Approach
Human Managed built an event-driven, near-real-time spine on AWS:
AWS Lambda, AWS Step Functions, and Amazon EventBridge for automated ingestion-through-response workflows.
Amazon SageMaker for ML used in dynamic threat detection and contextual risk scoring.
Amazon Managed Streaming for Apache Kafka (MSK) for streaming orchestration; Dagster for core coordination; Amazon EKS and Amazon EC2 for elastic compute; AWS Glue for lighter ETL.
AWS Enterprise Support is cited as tightening architecture and spend discipline — for example Savings Plan utilization rising from ~60% to ~98% on eligible workloads — so more engineering time lands on product rather than firefighting.
Forward-looking note in the same case study: Amazon Bedrock integration into a model context protocol (MCP) frame for enterprise generative agents (summarization, contextual query, cross-source analysis, compliance assistance).
Impact
Platform scale
> 3 billion events processed per day for near-real-time decisions.
> 1.2 PB of uncompressed security data ingested per month.
> 100,000 enterprise assets under monitoring for visibility, compliance, and health signals.
Response time
~97.9% lower mean time to respond (MTTR) for DDoS-class incidents.
~97.8% lower MTTR for phishing-class incidents.
Internal efficiency
~60% reduction in infrastructure-team workload — capacity redirected toward customer-facing innovation.
Customer outcome (bank example)
~90% reduction in network security violations when consolidating hybrid visibility with automated detection and response.
Incident response cadence cited as moving from months to minutes, with compliance obligations maintained across regional operations.
Karen Kim closes the published case study by tying infrastructure strength to speed, scale, and alignment with data-driven enterprises — the same thread as Human Managed’s purpose: empower responsible decisions under real constraints, not in theory.
Roshanara Basheerudin (Director) notes the shift from figuring out cloud alone to using depth of AWS support for optimization and go-to-market velocity — honest credit to partnership where it changed outcomes.
