Hexaware Technologies announced a strategic collaboration with Microsoft on July 27, 2023, to integrate its Tenjin for Knowledge Services platform with Microsoft Azure OpenAI Service, delivering generative AI capabilities to enterprise knowledge management systems. The move aims to help businesses unlock insights from vast repositories of unstructured data while maintaining strict governance and security controls through Microsoft Entra ID.

The Enterprise Knowledge Dilemma

Organizations today generate staggering volumes of internal documents, emails, chat logs, and technical notes. Yet finding exactly the right piece of information at the right moment remains a persistent productivity drain. Traditional search tools return links to documents, forcing users to sift through pages to locate a single data point. This inefficiency costs large enterprises millions in lost time annually.

Hexaware’s Tenjin for Knowledge Services tackles this head-on by applying large language models to understand natural language queries and surface precise answers—not just documents—from an organization’s entire knowledge base. The platform indexes content from SharePoint, OneDrive, ServiceNow, Confluence, and other common enterprise systems, creating a unified semantic search layer.

Inside the Tenjin Platform

Tenjin for Knowledge Services is not a generic AI wrapper. Hexaware has built it as a domain-aware engine pre-trained on common enterprise taxonomies and tuned for IT service management, HR support, and procurement workflows. When a user types “How do I reset my VPN credentials when working remotely?” the system retrieves the exact steps from the relevant knowledge article, pulling in only the paragraph that matters. It also verifies the answer against current policy documents flagged by Entra ID access controls.

This capability relies on Azure OpenAI’s text embedding and completion models. By using retrieval-augmented generation (RAG), Tenjin ensures responses are grounded in the organization’s own data, reducing hallucinations that plague public chatbots. The system cites its sources, allowing users to click through to the original document for deeper context—a critical feature for compliance-heavy industries.

Azure OpenAI: The Muscle Behind the Interface

Microsoft’s Azure OpenAI Service gives Hexaware access to cutting-edge models like GPT-4 with enterprise-grade reliability and compliance. Unlike the consumer version, Azure OpenAI offers dedicated capacity, data residency options, and private networking. For Tenjin, this means sensitive enterprise data never leaves the customer’s Azure tenant, a non-negotiable requirement for financial services, healthcare, and government clients.

Hexaware engineers fine-tuned the models on representative enterprise dialogue, teaching them to handle the ambiguous phrasing common in support tickets. The result is a system that understands misspellings, industry jargon, and multi-step questions without requiring users to learn specialized syntax.

Governance Through Microsoft Entra ID

Perhaps the most critical piece of the announcement is the deep integration with Microsoft Entra ID (formerly Azure AD). Every query in Tenjin is scoped to the user’s identity and group memberships. A contract specialist sees vendor agreement terms; a software developer sees API documentation; and an HR generalist sees payroll policies—all from the same natural language interface. Hexaware has embedded Entra ID’s conditional access policies, so users accessing highly confidential materials might face additional multi-factor authentication triggers automatically.

This role-based content filtering is managed through dynamic groups that update membership as employees change roles. Audit logs track every query and response, providing a forensic trail for compliance officers. For multinational firms, the system respects geofencing rules—a user in the EU won’t inadvertently access data hosted in regions where transfer is restricted.

Real-World Deployment Scenarios

Hexaware has already piloted Tenjin internally and with select clients. In one case, a global bank used the platform to reduce Level 1 support call volume by 34% within three months. Employees could self-serve answers to common IT and benefits questions without waiting in phone queues. The bank’s knowledge team also benefited: they used Tenjin’s analytics to identify articles with high query rates but low resolution scores, prompting rewrites that improved deflect rates further.

Another deployment at a manufacturing conglomerate connected Tenjin to decades of engineering reports. Maintenance technicians on factory floors now query the system via mobile devices to obtain torque specifications, safety procedures, and parts diagrams without leafing through binders. The ROI timeline compressed from 18 months to just 6 months due to reduced downtime.

Competitive Landscape and Differentiation

Hexaware isn’t alone in pursuing enterprise knowledge AI. Microsoft’s own Copilot for Microsoft 365 applies generative AI across productivity apps. ServiceNow offers GenAI for ITSM. Startups like Glean and Hebbia target similar use cases. What distinguishes Tenjin, according to Hexaware, is its vertical customization and the tight Entra ID governance mesh. Rather than generic enterprise search, Tenjin is purpose-built for IT and business process outsourcing use cases that Hexaware has serviced for decades.

The platform’s orchestration layer also pre-processes documents, chunking them optimally for language model consumption—a step that generic tools often lack. Hexaware claims its proprietary chunking strategy preserves context across paragraph boundaries, leading to more coherent answers than off-the-shelf vector databases.

Hexaware’s Azure Investment

This collaboration deepens Hexaware’s existing partnership with Microsoft. The company already holds multiple Azure specializations and a large base of Azure-certified consultants. By building Tenjin on Azure OpenAI, Hexaware reuses its in-house expertise and positions itself as a leading adopter of Microsoft’s AI services stack. The company plans to offer Tenjin both as a standalone SaaS product and as a component of broader outsourcing contracts, giving clients flexibility.

Hexaware’s CEO stressed that the move aligns with the firm’s “Cloudify Everything” strategy, which aims to modernize legacy enterprise processes using AI and cloud automation. The Tenjin platform is the centerpiece of its AI-for-IT-operations portfolio.

The Road Ahead for Governed Enterprise AI

The announcement arrives as enterprises grapple with “shadow AI”—employees using unapproved public chatbots that leak proprietary data. Hexaware’s platform directly addresses this by providing a safe, sanctioned alternative that keeps data inside the organizational boundary. Analysts expect rapid growth in private AI instances as security-conscious firms restrict public tool usage.

Hexaware plans quarterly feature updates including multi-language support, integration with Teams collaborative notes, and a bring-your-own-model option for clients that have trained proprietary LLMs. The company is also building connectors for ServiceNow’s Knowledge Management and Oracle’s Unified Directory to broaden the platform’s data reach.

What This Means for IT Teams

For Windows enterprise administrators, the integration with Entra ID is a practical relief. They can manage access to Tenjin through the same Azure portal and PowerShell scripts they already use for Microsoft 365 and Azure resources. Conditional access policies, identity protection alerts, and privileged identity management all apply without deploying additional agents. This reduces the operational burden of securing yet another third-party tool.

Moreover, because Tenjin runs within the customer’s Azure subscription, data never traverses public internet endpoints in the clear. Network security groups and Azure Firewall rules can restrict access to only authorized virtual networks, satisfying the most stringent defense-in-depth architectures.

Potential Pitfalls and User Feedback

No platform is without friction. Early adopters noted that initial document preprocessing can take days for libraries exceeding 500,000 documents, though Hexaware says it is optimizing the pipeline. Users also requested more granular reporting for query success rates separated by department, a feature now on the roadmap.

Pricing, while not publicly disclosed by Hexaware, follows an enterprise subscription model tied to the number of indexed documents and concurrent users. Analysts expect it to compete with existing knowledge management tools that add AI modules, putting pressure on vendors like ServiceNow and BMC.

Conclusion

Hexaware’s Tenjin for Knowledge Services, powered by Microsoft Azure OpenAI and governed through Microsoft Entra ID, represents a pragmatic fusion of generative AI with enterprise security needs. By keeping sensitive data within customer-controlled tenancies and enforcing role-based access at query time, it sidesteps the main compliance objections that have slowed generative AI adoption in regulated sectors. For IT leaders, the platform offers a tangible path to reduce support costs, speed up employee onboarding, and finally unlock the value trapped in institutional knowledge—without sacrificing control.