How DXC Technology built a multi-model AI substrate on Amazon Bedrock
DXC Technology runs mission-critical managed IT services for regulated organizations across 70 countries. Learn how DXC built DCX OASIS, its Human+AI operating model, on a governed multi-model substrate in Amazon Bedrock, separating a resilient model substrate from a trusted harness that orchestrates, governs, and audits AI agents at scale.
First seen by Oriel on .
AI summary
DXC Technology built DXC OASIS on Amazon Bedrock as a governed, multi-model substrate and trusted agent harness, reducing average resolution time by 95% across 56 measured IT operations use cases.
- Amazon Bedrock provides one API for Anthropic Claude, Amazon Nova, Meta Llama, and open-weight models, enabling routing by capability, latency, and cost.
- The private architecture runs agents on Amazon EKS and sends model calls through VPC endpoints and AWS PrivateLink, with encryption, tenant isolation, observability, and regional controls.
- The LangGraph-based harness adds orchestration, operational context, policy enforcement, four human-authority tiers, approval gates, and end-to-end action auditing beyond Amazon Bedrock Guardrails.
- DXC OASIS reached general availability in March 2026 and is scaling toward 130 customer environments using per-tenant namespaces and GitOps rollout.
- Agents support 12 operational personas, achieve at least 92% diagnostic accuracy, and reduced representative tasks from hours or days to minutes.
#amazon bedrock #ai agents #enterprise it #ai governance
Generated by gpt-5.6-sol. Check the original for details.
Aliases
| Source | External ID | URL |
|---|---|---|
| AWS Machine Learning Blog | 0e166aea5fc5e99de63f67eb0887a9fb579fb51f | https://aws.amazon.com/blogs/machine-learning/how-dxc-technology-built-a-multi-model-ai-substrate-on-amazon-bedrock/ Read discussion |