SelectHub has announced the launch of DataGrout, a new platform designed to help enterprises manage AI usage, optimize Large Language Model (LLM) costs, and enhance governance. The platform, developed by SelectHub's specialized AI research lab, aims to address the growing financial strain on corporate budgets caused by unoptimized LLM usage in enterprise settings.
DataGrout focuses on reducing token consumption for agentic workflows, chatbots, and various AI tools. It also provides IT and FinOps leadership with a policy-driven, auditable system for monitoring LLM payloads and associated costs, offering visibility into company-wide AI utilization. This is particularly relevant as enterprises increasingly adopt context-intensive chat sessions and coding tools like Claude Code, Cursor, and Replit, alongside expanding agentic workflow deployments in areas such as customer service and sales.
The platform functions as a unified connection layer, enabling AI agents to securely access enterprise applications. It incorporates dynamic context pruning and a symbolic inference layer to minimize calls to the underlying LLM. Early tests conducted by SelectHub indicate an average token reduction of 60% for data-intensive tasks involving integration with ERP and CRM systems, reportedly without compromising workload accuracy.
According to Venkat Devraj, CEO of SelectHub, DataGrout provides IT and finance teams with precise token tracking and insights into LLM consumption patterns and payloads. This allows leadership to measure the value of AI investments and systematically eliminate waste on a per-LLM-call basis. The platform is designed to keep deterministic workloads off the LLM entirely, without requiring users or developers to manually segregate and route these tasks.
Industry leaders have echoed the need for such tools. Deepak Dodani, CIO at AFS Logistics, highlighted the dual challenges of token economics and governance in AI deployment, noting that many companies struggle to assess whether LLM resources are driving business value or being consumed by unnecessary projects. Bharath Prabhakaran, Chief Digital Officer at the University of Cincinnati, also emphasized concerns among CIOs and CFOs regarding rapidly growing AI operational expenses due to rampant LLM token usage, making ROI forecasting difficult without clear usage and consumption metrics.
DataGrout operates as a Multi-Cloud Platform (MCP) gateway and control plane. Agents can connect via its Conduit SDK or through MCP or JSON-RPC. The platform supports access to any third-party MCP server and also offers its own MCP servers for popular SaaS applications, including Salesforce, HubSpot, Oracle ERP, SAP, ServiceNow, Snowflake, and QuickBooks. For LLM connectivity, DataGrout allows companies to use their own keys directly or connect via established AI gateways such as Amazon Bedrock and Kong.
