Merge integrates OpenAI technology to scale AI-powered business operations and product development

Overview of Merge's AI Integration

Merge has adopted OpenAI’s technology to power its suite of business and product tools, focusing on connecting software systems and enabling AI agents to execute actions. The company utilizes a multi-layered approach, ranging from data integration and secure tool access to model routing and cost management. By integrating these capabilities, Merge supports both the software products it builds for customers and the internal operational workflows that drive its business.

Technical Infrastructure and Reliability

Merge’s platform architecture is designed to prioritize reliability for AI-driven tool execution. According to CEO and Co-founder Shensi Ding, tool calls through Merge achieve a 99.9% success rate, which the company contrasts with public Model Context Protocol (MCP) servers that typically operate at approximately 60% reliability. The company’s core offerings include:

  • Merge Unified: Facilitates customer-data integrations.
  • Merge Agent Handler: Provides secure tool access for AI agents.
  • Merge Gateway: Manages model routing and cost controls.
  • Merge for Workforce: Enables internal teams to manage access to models, tools, and business systems.

AI-Driven Operational Efficiency

Merge has implemented AI-driven automation to streamline high-touch business processes, most notably in the planning and execution of executive dinners. By connecting GPT-6 Astra to systems like Salesforce, Gong, Google Sheets, and Gmail, the company has reduced the time required to manage guest lists, outreach, and attendee tracking from dozens of hours to approximately 30 minutes per event. This efficiency gain contributed to an increase in event frequency, moving from eight dinners in Q2 2026 to 16 in Q3, with plans for over 30 in Q4.

Product Development and Internal Adoption

Merge utilizes GPT-6 Astra and Codex to accelerate product development and internal productivity across non-engineering teams.

  • Code Development: The team uses Codex for code reviews, including the configuration interfaces for Merge for Workforce. Ding notes that Codex provides specific feedback, while engineers retain responsibility for validating and shipping the final implementation.
  • Cross-Functional Empowerment: Non-engineering staff use AI to build reusable workflows. For example, the sales leadership team uses AI to automate deal approvals and forecast updates, while the marketing and social media teams use GPT-6 Astra to refine messaging and monitor engagement.

Leadership Perspectives on AI Adoption

Merge’s leadership emphasizes a culture of experimentation and executive-led AI fluency. Key takeaways from their implementation strategy include:

  • Executive Modeling: Leadership believes executives must set the standard for AI fluency within an organization.
  • Low-Risk Experimentation: Initial AI testing was conducted on new, non-critical products to allow for learning without impacting core customer-facing codebases.
  • Autonomy: Employees are encouraged to choose their own AI tools and are given dedicated time to explore and test new capabilities, even if those experiments do not always yield immediate results.

Sources