Sold by: Arhasi, AI with Integrity
Deployed on AWS
TrustHouse.AI is a container-based trust and context layer for production-grade AI that runs in your AWS account. It applies AI governance policy controls, business context, and monitoring to AI agents and models. Available in three tiers, Essential, Production, and Autonomous, sized from 1 governed AI environment and 5 agents up to 10 environments and 100+ agents.
Overview
TrustHouse.AI is a unified trust and context layer for production-grade enterprise AI.
It deploys as containers inside your own AWS account and sits between your AI agents and the models they call, including models on Amazon Bedrock. Every request and response passes through TrustHouse, where your business policies are evaluated before anything reaches a model or a user.
Your data, models, and business logic stay inside your AWS boundary.
Every tier includes the same core capabilities.
- The AI Risk Engine detects, validates, and enforces business policies on AI inputs and outputs, with actions such as block, mask, or a standard fallback response.
- The Secure AI Gateway and MCP Proxy route agent traffic through those policies.
- Zero Trust for AI Agents
- Shadow AI and Access Control govern which users and agents can reach which models and tools.
- Agent Discovery and Catalog gives you an inventory of the agents running in your environment.
- Continuous Assurance Monitoring tracks agent behavior over time.
- Context Engine grounds models in your business vocabulary, taxonomy, and ontology to reduce hallucinations.
- AgentOps supports the build, deployment, security scanning, and runtime testing of agents.
TrustHouse is available in 3 tiers on the same trust foundation, so you can change tiers with no re-architecture:
-
Essential is for companies beginning to operationalize AI: 1 governed AI environment or account, up to 50 users, 1 AI model or platform, 3 AI applications or workloads, and 5 agents, with standard custom policies and controls, quarterly trust reporting, business-hours incident response, and an annual architecture review.
-
Production is for companies running AI in production: 3 governed AI environments or accounts, up to 200 users, 3 AI models or platforms, 15 AI applications or workloads, and 25 agents, with advanced custom policies and controls, monthly trust reporting, priority incident response, a quarterly architecture review, and dedicated Trust engineering.
-
Autonomous is for organizations where AI is business-critical: 10 governed AI environments or accounts, up to 500 users, 6 AI models or platforms, 50+ AI applications or workloads, and 100+ agents, with advanced custom policies and controls, monthly and executive trust reporting, 24x7 incident response, a quarterly architecture review, and dedicated Trust engineering.
Highlights
- Runs in your AWS account. TrustHouse deploys as containers inside your VPC, so prompts, responses, and business data stay within your AWS boundary while AI governance policies are enforced on every request.
- One trust layer for AI agents and models: AI Risk Engine, Secure AI Gateway and MCP Proxy, Zero Trust for AI Agents, Shadow AI and Access Control, Agent Discovery and Catalog, Continuous Assurance Monitoring, Context Engine, and AgentOps.
- 3 tiers on one trust foundation: Essential (1 environment, 5 agents), Production (3 environments, 25 agents), and Autonomous (10 environments, 100+ agents). Change tiers as your AI footprint grows.
Details
Sold by
Categories
Delivery method
Supported services
Delivery option
TrustHouse on Amazon ECS (Arhasi-assisted deployment)
Latest version
Operating system
Linux
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
Custom pricing options
Pricing is based on your specific requirements and eligibility. Request a private offer to receive a custom quote. Sign in to view any offers that have been extended to you.
Legal
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
TrustHouse on Amazon ECS (Arhasi-assisted deployment)
Supported services: Learn more
- Amazon ECS
Container image
Containers are lightweight, portable execution environments that wrap server application software in a filesystem that includes everything it needs to run. Container applications run on supported container runtimes and orchestration services, such as Amazon Elastic Container Service (Amazon ECS) or Amazon Elastic Kubernetes Service (Amazon EKS). Both eliminate the need for you to install and operate your own container orchestration software by managing and scheduling containers on a scalable cluster of virtual machines.
Version release notes
TrustHouse 2.0.0 (initial release on AWS Marketplace) Update label: Important. This is the first version, so there is nothing to upgrade from.
First release of TrustHouse.AI on AWS Marketplace: the complete platform as 48 container images for Amazon ECS on AWS Fargate, deployed by Arhasi into your own AWS account. Container image tag: trusthouse-v2.0.2-mp2.
Included:
- The trust layer: AI Risk Engine, Secure AI Gateway and MCP Proxy, Zero Trust for AI Agents, Shadow AI and Access Control, Agent Discovery and Catalog, Continuous Assurance Monitoring, Context Engine and AgentOps.
- Data Catalog, Data Lineage, and Data Quality and Data Profile.
- Built-in AI features use Amazon Bedrock in your own AWS account.
New for AWS Marketplace:
- Licensing through AWS License Manager in your account, with no licence key.
- No AWS access keys: TrustHouse reaches your AWS data and Bedrock agents only through the TrustHouseAccess IAM role you create from its CloudFormation template. Its own services use IAM task roles in the account where it runs.
- AgentOps builds and runs your own agents in your account with AWS CodeBuild and AWS Fargate. Agent images are scanned by Amazon ECR.
- Every image runs as a non-root user and was scanned by Arhasi for vulnerabilities and secrets before publishing.
Additional details
Usage instructions
TrustHouse runs in your AWS account on Amazon ECS (Fargate), deployed by Arhasi from this version's 48 images: 709825985650.dkr.ecr.us-east-1.amazonaws.com/arhasi-ai-with-integrity/trusthouse-<service>:trusthouse-v2.0.2-mp2 Deployment guide ("guide N" = its section N): https://dv2fy2582x7m5.cloudfront.net/TrustHouse_Deployment_Guide.pdf
WHAT YOU DO
- Subscribe in the AWS account where TrustHouse will run.
- Email contactus@trusthouse.ai your AWS account ID and Region.
- Create a stack from Arhasi's CloudFormation template with the External ID Arhasi sends separately: one IAM role, ArhasiDeployer, for Arhasi's deployment account only. Arhasi's first deploy adds the CDK bootstrap stack (execution role with AdministratorAccess). Delete the stack to end Arhasi's access (guide 5).
- License Manager console, your Region: Start using License Manager (first time); if the TrustHouse licence is not activated, Accept and activate it (guide 4).
- As Arhasi advises: quota increases (guide 13) and the Bedrock first-use form for Anthropic models.
- After handover: sign in at your TrustHouse address with Google or Microsoft, invite your users, confirm the alarm email subscription.
- In TrustHouse, Integrations > AWS role: create the TrustHouseAccess role from its template; paste its RoleArn into each AWS connection. No AWS keys; you decide what it may reach (guide 6).
WHAT ARHASI DOES Checks access and Bedrock models through the role; sets up domain and certificate, sign-in, Bitbucket for AgentOps; deploys ECS services, DocumentDB, ElastiCache, load balancer and S3 in a new VPC, and CodeBuild and ECR for AgentOps.
COSTS AND QUOTAS (guide 13, 14) AWS bills your account. Full install, us-east-1 list prices: about USD 1,146 a month, USD 34 a day idle. With fewer capabilities it costs less; Arhasi gives you an estimate for your install before deploying. Plus Bedrock tokens, Container Insights, agents and data transfer. Quotas: Fargate 21 vCPU, 42 during upgrades, plus 10 for agents; 2 CodeBuild builds at once; 2 more VPCs.
LICENCE CHECK AWS License Manager in your account, at start-up and every 5 minutes; no licence key or outside server. Unreachable: sign-in works up to 72 hours. No active subscription: sign-in refused.
SENSITIVE DATA (guide 10) Passwords, keys and connection credentials are in AWS Secrets Manager, or in DocumentDB encrypted with TrustHouse's application key (kept in Secrets Manager). Encryption: DocumentDB and ElastiCache at rest and TLS; S3 server-side; Secrets Manager AWS managed keys. Never change or delete the application key alone: saved credentials become unreadable. DocumentDB password rotates every 30 days; other keys with Arhasi support.
EXTERNAL CONNECTIONS (guide 9) Ongoing: Google or Microsoft sign-in. Other connections depend on what your install includes; guide 9 lists them. No TrustHouse container image comes from outside AWS Marketplace; the deploy also adds one TrustHouse Lambda function and CDK helpers. Optional:
- Your own data sources and code repositories, with your own logins.
- AgentOps builds your own agent code with CodeBuild, downloading base images from Amazon ECR Public and packages (PyPI, npm, Maven, Debian). Agents call out on ports 80 and 443; a published agent accepts HTTP on 8080 from the internet. Images scanned by Amazon ECR (basic or Inspector).
- Veracode: off by default; only with your own Veracode account and licence. When on, agent source code goes to Veracode.
HEALTH (guide 12) ECS console, cluster arhasi-<environment>-cluster: running = desired for every service. Logs: CloudWatch Logs. Alarms: SNS topic arhasi-<environment>-alarms.
Support: contactus@trusthouse.ai
Resources
Support
Vendor support
TrustHouse support is provided by Arhasi Inc.
Email: contactus@trusthouse.ai
Web: trusthouse.ai
Support includes deployment assistance, configuration guidance, and issue resolution for the TrustHouse containers running in your AWS account. Incident response by tier: business hours (Essential), priority (Production), 24x7 (Autonomous).
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
Atlan is the context layer for AI, the infrastructure that makes enterprise AI accurate, trustworthy, and scalable. Trusted by Mastercard, Workday, General Motors, and hundreds of global enterprises, Atlan connects your entire data estate and enriches it with the certified definitions, lineage, ownership, and access policies every AI agent needs. Native integrations with Amazon Redshift, S3, Glue, Athena, Lake Formation, and SageMaker Unified Studio mean your AWS stack is context-ready from day one. Recognized as a Leader in the Gartner Magic Quadrant for Data & Analytics Governance 2026 and Customer Favorite in the Forrester Wave for Data Governance Q3 2025.
Sphinx AI is a NYC-based, research focused startup working on training algorithms to learn and maintain the institutional knowledge that typically lives in employees' heads or is embedded across an enterprise's dashboards, code, presentations, documentation, notes, and other work products. Rather than relying on manual prompt engineering to maintain context, we automatically enable AI to understand how your business defines, uses, and reasons about data. We're working with enterprise data teams to improve the accuracy, auditability, and ROI of AI on internal data. We've helped marketing teams self-serve complex modeling on proprietary data, enabled operations teams to solve supply chain problems without a data scientist in the loop, and helped data teams themselves deploy autonomous agents that don't need constant babysitting.We also reduce the operational burden of keeping AI context up to date, saving thousands of hours a year even for relatively small data teams.
Tacnode Context Lake™ transforms live data into real-time context for AI. Learn the full definition in our Context Lake manifesto.
NeuBird delivers The Production Ops Agent (Prod Ops Agent), powered by the Agentic Reliability Center (ARC), to autonomously prevent, resolve, and operate modern production. Going far beyond narrow AI SRE tools that only react after an alert fires, the Prod Ops Agent closes the loop across the entire operations lifecycle. It catches silent drift and degradations 30 to 60 minutes before alerts fire, accelerates Incident Resolution, and helps Reduce MTTR through autonomous RCA, correlating 15+ sources in parallel across AWS, Datadog, Splunk, and GitHub to isolate root cause in under 5 minutes at 94% accuracy. It then executes fixes via configurable human approval gates. Beneath the agent, ARC queries 50+ tools in place with zero telemetry storage, cutting LLM token spend significantly. It compounds institutional memory and serves governed context to human teams and custom agents/IDEs via MCP, eliminating 80% of P1 war rooms and reclaiming 40% of engineering capacity.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.