Enterprise Guide
This guide is for enterprise data teams evaluating and deploying Data Workers.
Deployment Options
Data Workers supports three deployment models to match your security and compliance requirements:
SaaS (Multi-Tenant) — Managed service with no infrastructure to maintain. Data is logically isolated between tenants. Fastest path to production.
Single-Tenant Cloud — Dedicated instance deployed in your AWS, GCP, or Azure account. You control the infrastructure, we manage the application. Full network isolation.
VPC / On-Premise — Data Workers runs entirely within your network perimeter. No data leaves your environment. Suitable for regulated industries and air-gapped deployments. VPC and on-premise deployments ensure data never leaves your chosen region.
Security Overview
Data Workers implements zero-trust architecture with encryption at rest and in transit, automated PII scrubbing, and immutable audit trails. See the Security & Compliance doc for full details.
Compliance
Data Workers is architected to meet SOC 2 Type II and HIPAA requirements. Contact us for our current security assessment documentation.
- •7-year audit log retention
- •Continuous dependency monitoring with quarterly security assessments
- •Annual penetration testing
- •PII detection middleware — automatically detects and scrubs sensitive data before LLM processing
- •Promotion audit trails — full traceability for every change promoted across environments
Community vs Enterprise
Community: All 15 agents, self-hosted deployment, community support (GitHub), basic audit logs, build your own custom agents.
Enterprise: All 15 agents + custom, SaaS/VPC/On-prem deployment, dedicated support with SLAs, SSO (Okta, Azure AD), SOC 2 & HIPAA architecture, advanced audit logs (7-year retention), custom agents with support, Medallion architecture support (Bronze→Silver→Gold autonomous management), multi-provider LLM support (bring your own Claude/OpenAI/Bedrock/Vertex keys), usage metering and cost tracking.
SLAs
Enterprise customers receive SLAs tailored to their deployment model:
- •P1 — Critical: Platform down, agents not running. Target response: 1 hour. Target resolution: 4 hours.
- •P2 — High: Agent degraded, partial functionality loss. Target response: 4 hours. Target resolution: 1 business day.
- •P3 — Medium: Non-critical issue, workaround available. Target response: 1 business day. Target resolution: 3 business days.
- •P4 — Low: Feature request, documentation, minor issue. Target response: 2 business days. Target resolution: Best effort.
Uptime target: 99.9% on a rolling 30-day window for SaaS and single-tenant deployments.
Getting Started for Enterprises
We recommend a phased rollout:
Phase 1: Read-Only Agents — Start with agents that observe and report without making changes: Incident Debugging Agent and Quality Monitoring Agent. These agents provide immediate value with zero risk to your production environment.
Phase 2: Configure Autonomy and Approvals — Once you are comfortable with agent behavior, set autonomy levels per agent, configure approval workflows (Slack, email, dashboard), and define confidence thresholds for auto-execution vs. escalation.
Phase 3: Expand to Action-Taking Agents — Gradually enable agents that can make changes: Schema Evolution Agent, Pipeline Building Agent, and Cost Savings Agent. All action-taking operations go through your configured approval gates.
Phase 4: Monitor with Usage Intelligence — Enable the Usage Intelligence Agent to see how your team uses the platform. Track tool usage patterns, identify workflow bottlenecks, measure adoption, and maintain full audit trails for compliance.
Support
Community — GitHub Issues for bug reports and feature requests. Community discussions for general questions.
Enterprise — Dedicated support channels with SLAs, onboarding assistance and deployment support, quarterly business reviews.
Pricing — Contact hello@dataworkers.io for enterprise pricing.