DXC Partnered with LangChain for Governed AI Agents
The integration aims to scale agentic IT operations with human-in-the-loop oversight.
Updated on Oct. 13, 2026 in Artificial Intelligence
Live Poll
Do you trust businesses to use AI agents for managing critical infrastructure?
DXC Technology has partnered with LangChain to incorporate the LangGraph and LangSmith frameworks into its existing OASIS platform. The production-ready system is currently scaling to support 125 enterprise customer environments.
Why it matters
This partnership targets the challenge of enterprise AI adoption by embedding governance and human oversight directly into automated IT workflows. It seeks to bridge the gap between experimental agentic capabilities and reliable, accountable production systems.
The OASIS platform reduced filesystem incident resolution from 53-90 minutes to 3-6 minutes, and database incident resolution from 2-3 hours to 5-12 minutes. The system is designed to manage 500 distinct operational use cases.
The players
DXC Technology
An IT services company that provides enterprise infrastructure, cloud migration, and managed services for large organizations.
LangChain
A software provider that develops frameworks for building LLM-powered applications and agentic orchestration tools.
The details
DXC utilizes LangGraph — a library for building stateful, multi-actor applications with LLMs — to orchestrate agent workflows. LangSmith is employed for development, monitoring, and evaluation of agent performance. The platform integrates human-in-the-loop workflows, which act as a verification gate allowing human experts to approve sensitive agent actions before execution.
Timeline
October 13, 2026: DXC Technology and LangChain announced the partnership.
The Tech Race
The integration marks a significant attempt to standardize agentic AI within established IT service delivery models. It follows a growing industry trend of moving from simple chatbot interfaces to complex, orchestrated agents that handle multi-step operational tasks.
Enterprise IT teams using the OASIS platform gain access to accelerated incident resolution workflows that handle complex problem management in minutes rather than days. These capabilities are currently live and scaling toward 125 production environments for DXC clients.
The takeaway
The move demonstrates that the path to enterprise-grade AI involves rigorous observability and human-approval gates rather than just raw model performance. Watch for performance stability data as the platform scales toward its target of 125 customer environments.
Further reading
For more on the development of intelligent automation systems, see the latest research in Artificial Intelligence.
Source note: This article includes information reported by The Kingston Whig-Standard.
Live Poll
Do you trust businesses to use AI agents for managing critical infrastructure?
