DXC and LangChain Advance Agentic AI Harnesses for Mission-Critical IT Operations

DXC’s OASIS platform cuts IT incident resolution times by up to 95% using AI agents.

What the company is saying

DXC Technology is highlighting a strategic partnership with LangChain to advance governed agentic AI in enterprise IT, positioning itself as a leader in operationalizing AI agents for mission-critical environments. The company emphasizes that OASIS, its orchestration platform built on LangChain’s LangGraph and supported by LangSmith, is already in production and scaling toward 125 customer environments. DXC claims OASIS delivers measurable operational improvements, specifically a 95% average reduction in resolution time and diagnostic accuracy of 92% or better across 500 use cases, with validation against senior engineer judgment. The announcement stresses governance, human oversight, and flexibility, stating that OASIS integrates agent actions with controls for reversibility, impact, and human approval when needed. Dan Gray, President of Global Infrastructure Services at DXC, and Julia Schottenstein, COO of LangChain, are quoted to reinforce the credibility and technical depth of the partnership. The company also points to a new technical white paper detailing its approach to AI agent governance and oversight.

What the data suggests

The disclosed numbers show OASIS delivering substantial operational gains: across 500 measured use cases, average resolution time for incidents is down 95%, and diagnostic accuracy is at least 92% when checked against senior engineers. Specific examples include reducing filesystem incident resolution from 53–90 minutes to 3–6 minutes, database performance incidents from 2–3 hours to 5–12 minutes, and complex problem management from 2–6 days to 8–10 minutes. The platform is not yet at the 125-customer target but is described as scaling toward that milestone. While operational metrics are concrete and validated, there is no financial data—such as revenue impact, cost savings, or customer adoption rates—so the business impact remains unquantified. Claims about model agnosticism, interoperability, and governance are asserted but not supported by quantitative evidence or third-party validation. The technical and operational disclosures are specific, but the absence of financial context limits assessment of overall business value.

Analysis

The announcement is generally positive in tone and provides specific, measurable operational improvements (e.g., 95% reduction in resolution time, 92%+ diagnostic accuracy, and concrete before/after incident resolution times). These realised metrics support the claim that the OASIS platform is delivering value in production environments. However, the narrative inflates the signal by emphasizing scaling ambitions ('scaling toward 125 customer environments') and broad technical/governance capabilities without providing evidence for current adoption levels, financial impact, or customer-specific outcomes. The majority of key claims are realised, but the most forward-looking statements relate to future scaling and continuous improvement, which are not yet substantiated by disclosed data. No large capital outlay is disclosed, and the operational improvements are already being realised, so capital intensity is not a concern. The absence of any financial metrics (revenue, profit, cost savings) means the true_signal cannot exceed weak_positive, as investors cannot assess the business impact or sustainability of these operational gains.

Risk flags

• ●There is no disclosure of financial impact—no revenue, cost savings, or customer adoption figures—so investors cannot assess whether operational gains translate to business value or profitability.
• ●The claim of scaling toward 125 customer environments is forward-looking, with no current adoption numbers or customer names provided; actual market uptake may lag expectations.
• ●Technical and governance claims (model agnosticism, interoperability, human oversight) are asserted without quantitative validation or independent audit, leaving open the risk that real-world complexity or compliance requirements could limit deployment or effectiveness.
• ●The platform’s success depends on integration with diverse enterprise IT environments; unforeseen technical, regulatory, or change-management challenges could slow adoption or dilute results.

Bottom line

DXC Technology’s partnership with LangChain and the deployment of OASIS demonstrate that AI agents can deliver dramatic operational efficiencies in IT incident management, with resolution times cut by up to 95% and diagnostic accuracy exceeding 92% across hundreds of use cases. These are realised, not hypothetical, improvements, and the platform is already in production, though the 125-customer target remains a forward-looking goal. The lack of financial data means investors cannot yet judge whether these operational gains drive revenue, margin expansion, or competitive advantage. The narrative is credible on technical delivery but aspirational on market scaling and business impact. For investors, the most important next step is evidence of customer adoption and financial outcomes. The key takeaway: OASIS is delivering on operational metrics, but the business case still needs to be proven with hard financial results.

Announcement summary

(NYSE:DXC) DXC Technology announced a partnership with LangChain to advance governed agentic AI across mission-critical enterprise IT environments. The partnership leverages DXC's use of LangGraph within DXC OASIS, an intelligent orchestration platform and operating model for mission-critical IT operations. DXC is licensing LangSmith to support the development, evaluation, and observability of OASIS agents. OASIS is currently in production and scaling toward 125 customer environments across regulated industries and multiple geographies. The platform integrates LangGraph's open-source agent orchestration framework and LangSmith's tooling for agent development, evaluation, and observability. OASIS enables agents to respond to operational events, investigate issues in parallel, utilize specialized capabilities across technology domains, and involve human experts when judgment or approval is required. Across more than 500 measured use cases, OASIS has reduced resolution time by 95% on average, with diagnostic accuracy of 92% or better when validated against senior engineer judgment. Specific production use cases include reducing filesystem incident resolution from 53-90 minutes to 3-6 minutes, database performance incidents from 2-3 hours to 5-12 minutes, and complex problem management from 2-6 days to 8-10 minutes. The partnership focuses on expanding the use of AI agents while maintaining governance and human oversight, with OASIS enabling agents to operate within defined authority and controls required for mission-critical environments. Actions by agents are assessed for reversibility, impact, verifiability, and track record, with sensitive actions paused for human approval and all decisions, tool calls, and outcomes traced for visibility. OASIS grounds agents in specific enterprise environments, drawing on system-specific and customer-specific information, DXC operational expertise, and vendor documentation. The platform is model agnostic, supports interoperability with third-party and customer-owned systems, and allows organizations to select models based on quality, cost, latency, and privacy requirements. OASIS is designed for continuous improvement, using insights from agent executions to enhance performance and incorporate lessons from individual scenarios. Deployments can span public cloud, private cloud, dedicated, and sovereign environments without redesigning the underlying architecture. Customers retain control of their data, workflows, and operational capabilities, preserving flexibility across models, infrastructure, and deployment environments. DXC and LangChain have published a technical white paper, 'DXC OASIS: The Harness for Mission-Critical IT Operations,' detailing their approach to operationalizing AI agents with governance, accountability, and human oversight.

Disagree with this article?