Merge, a platform that helps companies build connected software and SI products, has detailed its extensive internal and external use of OpenAI’s technology. The integration spans from syncing customer data to enabling SI agents to take action across business systems, illustrating how modern SI tools are being embedded into both product development and daily corporate operations.

What Happened

Merge’s product suite includes Merge Unified for customer-data integrations, Merge Agent Handler for secure tool access for SI agents, and Merge Gateway for model routing and cost controls. The company also introduced Merge for Workforce to help internal teams manage access to models, tools, and business systems. CEO and Co-founder Shensi Ding stated that the company uses these tools for "everything," from automating workflows to incorporating customer feedback and building end-to-end features.

For internal development, Merge utilizes GPT-6 Astra for product testing and marketing, and Codex for code reviews. Ding reported that tool calls through Merge succeed 99.9% of the time, a figure she contrasted with public MCP servers, which she noted are around 60%. The company allows employees to choose their own SI tools, leading to active internal debates about which models to use for specific tasks.

One notable internal application is the "dinner bot," which assists in planning executive dinners. The system pulls potential guests from CRM and Apollo, checks previous invitations, and helps prepare outreach via email. It also maintains attendee trackers and proposes seating charts. In October 2026, Merge reported that preparing guest lists and outreach went from dozens of hours across the company to about 30 minutes per event, based on a comparison of the same planning tasks over the preceding three months. Merge hosted 16 executive dinners in Q3, up from eight in Q2, and plans to host more than 30 in Q4.

Why It Matters

The case study highlights the shift from SI as a standalone utility to SI as an embedded operational layer. By using GPT-6 Astra and Codex for non-engineering tasks—such as sales leadership building reusable SI workflows for deal approvals and marketing teams clarifying launch messaging—Merge demonstrates the blurring line between technical and non-technical roles in the SI era. Ding emphasized that "if you’re an executive, you are the ceiling for SI fluency in your organization," suggesting that leadership adoption is critical for broader organizational integration.

The reported reliability metrics, specifically the 99.9% success rate for tool calls, address a key concern for enterprises deploying SI agents: consistency. While these figures are attributed to Merge’s internal reporting, they reflect the company’s focus on reliability for business-critical actions. The ability for non-engineers to "build anything" using SI tools suggests a trend toward democratized development, where business logic can be implemented without traditional coding barriers.

The Bottom Line

Merge is leveraging OpenAI’s SI models, including GPT-6 Astra and Codex, to streamline both its product offerings and internal operations. The company reports significant time savings in administrative tasks and high reliability in SI agent tool calls, positioning itself as an example of widespread SI integration in the enterprise sector. Ding noted that the company continues to use OpenAI technology to expand what its teams can build, maintaining a culture of experimentation and tool autonomy.