👋
About
Learn more about autonomous cloud management and how to optimize for cost and performance.
Sedai is an autonomous cloud management platform powered by AI/ML delivering continuous optimization for cloud operations teams to maximize cloud cost savings, performance and availability at scale.
Sedai enables teams to shift from static rules and threshold-based automation to modern ML-based autonomous operations. Sedai independently detects, prioritizes, and analyzes data to identify opportunities to safely act in production as well as provide deep contextual performance insights. Through autonomous actions, Sedai continuously learns from production behavior to evolve its intelligence models by evaluating outcomes and improving its symptom detection.
Using Sedai, organizations can reduce cloud cost by up to 50%, improve performance by up to 75% and multiply SRE productivity by up to 6X. Sedai can perform work equivalent to a team of cloud engineers working behind the scenes to optimize resources and remediate issues, so organizations can focus back on innovation.
Sedai requires access to cloud environments to discover topology as well as a source for monitoring data to analyze performance behavior:
- Amazon Web Services (AWS), including:
- Lambda Serverless Functions
- Elastic Container Service (ECS)
- Fargate
- Elastic Compute Cloud (EC2)
- Elastic Block Store (EBS)
- Elastic File System (EFS)
- Simple Storage Service (S3)
- Elastic Kubernetes Service (EKS)
- Google Cloud Provider, including:
- Dataflow BETA
- Google Kubernetes Engine (GKE)
- Microsoft Azure, including
- Azure Kubernetes Service (AKS)
- Azure VMs
- Self-managed or hosted Kubernetes clusters, including:
- Stateless workloads (autonomous management)
- Stateful workloads (recommendations only)
- AppDynamics
- CloudWatch (by default, Sedai automatically connects to CloudWatch when you connect an AWS account)
- Datadog
- Netdata
- New Relic
- Prometheus
- Splunk (SignalFX)
- Wavefront (VMware)
Email [email protected] to request support for additional cloud resource types or monitoring providers.
Select a resource type to learn more about Sedai's unique value props:
Amazon Web Services
Lambda
ECS & Fargate
Storage
EC2
EKS
- Autonomous Optimization: Optimize cost and/or performance based on machine learning to meet your cost and performance goals.
- Autonomous Concurrency: Virtually eliminate cold starts with ML-based selection of the volume and mix of provisioned concurrency and warmups informed by traffic seasonality, without cost increases.
- Autonomous Remediation: Detect and remediate performance issues including timeouts and OOM.
- Release Intelligence: Get production performance analysis of every release with scorecards and insights on duration, cost, and errors changes.
- Service Level Objectives (SLOs): Track targets for autonomously defined p95 and p99 error and latency SLOs.
- Service Optimization: Configures horizontal & vertical scaling for the best cost & performance, optimizing memory & CPU.
- Container Instance Optimization: Selects instance types on an application-aware basis, factoring in application level latency needs.
- Autonomous Remediation: Detect and remediate performance issues including out of memory and restarts to eliminate Failed Customer Interactions (FCIs).
- Release Intelligence: Get production performance analysis of every release with scorecards and insights on latency, cost, and errors changes.
- Service Level Objectives (SLOs): Track targets for autonomously defined p95 and p99 error and latency SLOs.
- Optimization: Reduce expenses based on machine learning to meet your cost goals for EFS, EBS, and S3 resources
- Optimization: Reduce expenses based on machine learning to meet your cost goals for individual EC2 instances or instances grouped together by tags or load balancers.
- Autonomous Workload Optimization: Optimize horizontal and vertical scaling for the best cost and performance while adjusting memory, CPU and replica sets at the container and pod/task level.
- Autonomous Node Optimization: Select instance types on an application-aware basis, factoring in application level latency needs.
- Autonomous Remediation: Detect and remediate performance issues including out of memory and restarts to eliminate Failed Customer Interactions (FCIs)
- Release Intelligence: Get production performance analysis of every release with scorecards and insights on latency, cost, and errors.
- Service Level Objectives (SLOs): Track targets for autonomously defined p95 and p99 error and latency SLOs
Google Cloud Provider
Dataflow
GKE
- Optimization: Reduce expenses based on machine learning to meet your cost goals for your Dataflow resources
- Autonomous Workload Optimization: Optimize horizontal and vertical scaling for the best cost and performance while adjusting memory, CPU and replica sets at the container and pod/task level.
- Autonomous Node Optimization: Select instance types on an application-aware basis, factoring in application level latency needs.
- Autonomous Remediation: Detect and remediate performance issues including out of memory and restarts to eliminate Failed Customer Interactions (FCIs)
- Release Intelligence: Get production performance analysis of every release with scorecards and insights on latency, cost, and errors.
- Service Level Objectives (SLOs): Track targets for autonomously defined p95 and p99 error and latency SLOs
Microsoft Azure
AKS
VMs
- Autonomous Workload Optimization: Optimize horizontal and vertical scaling for the best cost and performance while adjusting memory, CPU and replica sets at the container and pod/task level.
- Autonomous Node Optimization: Select instance types on an application-aware basis, factoring in application level latency needs.
- Autonomous Remediation: Detect and remediate performance issues including out of memory and restarts to eliminate Failed Customer Interactions (FCIs)
- Release Intelligence: Get production performance analysis of every release with scorecards and insights on latency, cost, and errors.
- Service Level Objectives (SLOs): Track targets for autonomously defined p95 and p99 error and latency SLOs
- Optimization: Reduce expenses based on machine learning to meet your cost goals for individual Azure VMs or VMs grouped together by tags or load balancers.
Kubernetes
Kuberenetes
- Autonomous Workload Optimization: Optimize horizontal and vertical scaling for the best cost and performance while adjusting memory, CPU and replica sets at the container and pod/task level.
- Autonomous Node Optimization: Select instance types on an application-aware basis, factoring in application level latency needs.
- Autonomous Remediation: Detect and remediate performance issues including out of memory and restarts to eliminate Failed Customer Interactions (FCIs)
- Release Intelligence: Get production performance analysis of every release with scorecards and insights on latency, cost, and errors.
- Service Level Objectives (SLOs): Track targets for autonomously defined p95 and p99 error and latency SLOs
- To connect an AWS Account, you will need sufficient Identity and Access Management (IAM) permissions within your organization's AWS Console to complete the following:
- Attach IAM Policy to Sedai IAM Role or User
- Generate Role ARN or Access/Secret Keys
- To connect a Kubernetes cluster (either self-managed or hosted), you will need sufficient permissions within the cluster to deploy Sedai's Smart Agent as well as credentials to connect to your monitoring provider. Learn more about specific monitoring provider requirements here.
If you have questions about setting up your account or how to use Sedai, feel free to reach out:
- Discover: Sedai securely connects to cloud environments to understand topology and to monitoring providers to automatically identify and prioritize metrics. For first-time connections, the discovery process usually takes at least a half hour or longer depending on the size of the cloud environment.
- Analyze: Once connected to a cloud and monitoring source, Sedai takes about 14 days to understand initial performance trends and build a baseline for predictive analytics. The system intelligently analyzes monitoring metrics, traffic patterns, and resource dependencies to deeply understand performance in production. By overlaying this data, Sedai independently determines expected resource performance on a granular seasonality-based level. This in-depth analysis informs its decision models so that Sedai can detect unhealthy symptoms early on without manual thresholds.
- Acts: Sedai leverages its in-depth analysis of performance behavior to identify opportunities to reduce cloud cost, improve latency, and prevent availability issues. Sedai only generates autonomous actions when it has sufficient confidence and can guarantee safe execution within production.
- Learn: Sedai continuously learns from its actions, seasonality trends, and performance behavior to evolve its intelligence models.
Sedai works best in production environments that experience consistent traffic and usage so that it can continuously learn from performance behavior. If you're not sure if your environment has sufficient performance data, we recommend emailing [email protected] to discuss your goals with Sedai.
Last modified 18d ago