Make Reveals Internal AI Agent Results: 14% Faster Hiring, 98% Documentation Coverage.

Make has published results from its internal AI transformation program, showing how the automation platform used its own technology to solve problems across sales, product, and recruiting. The company ran 23 AI and automation projects in 2025 under a program called Miriwa, with three delivering measurable impact on revenue, documentation coverage, and hiring speed.

The announcement, shared by Make's leadership on LinkedIn, offers concrete numbers on what AI adoption looks like when implemented internally. The three projects tackled different operational problems but followed the same framework: identify structured work that agents can handle, define success criteria, and build governance before deployment.

Make's sales team had a CRM data problem. When reps logged why deals were lost, the recorded reason rarely matched reality. Before the agent system launched, 79% of lost deals had no loss reason at all. Of the 21% that did, 86% were wrong.

The Deal Loss Agent uses a multi-agent system to analyze sales call transcripts, identify the actual reason a deal fell through, and update the CRM record automatically. The system runs after calls conclude, cross-referencing what was said against loss categories and correcting entries that don't align with the conversation. Make didn't share adoption rates or accuracy metrics for the agent itself, but eliminating the manual logging step removes a common source of friction in sales workflows.

Make maintains over 1,200 community-built integrations. A significant portion had no documentation because the company lacked the capacity to write and maintain it manually. The Apps Docs Agent combines workflows with an AI agent to research third-party APIs, generate integration guides, and validate the output before publishing.

After deploying the system, Make reached 98% page coverage for integration documentation. The hybrid approach, using both structured workflows and an agent layer, handles the research and drafting while keeping a validation step to catch errors before content goes live. Documentation was previously a manual bottleneck that scaled poorly with the number of integrations.

Make's recruiting team was manually screening CVs and losing visibility into where candidates were in the hiring pipeline. The Recruiter Agent prequalifies applications, triages them based on role requirements, and maintains a clean overview of the hiring process. Time to hire dropped 14%, and the system surfaced an additional ten high-match profiles per role that would have been overlooked in manual review.

The agent-powered interface handles the repetitive work of initial screening, freeing recruiters to focus on later-stage evaluation and candidate experience. Make didn't specify whether the agent integrates with existing applicant tracking systems or operates as a standalone layer.

The Miriwa program's results highlight a pattern: AI agents work best when applied to structured, repeatable tasks where success can be defined clearly. Deal loss analysis, documentation generation, and resume screening all fit that profile. Make's approach, building governance and validation into each system before deployment, reflects lessons the company has shared publicly about building production-ready AI agents.

The company didn't disclose how many of the other 20 Miriwa projects are in production or what problems they addressed. The three highlighted projects represent the highest-value outcomes from the program, but the broader initiative suggests Make is using its own platform extensively for internal automation.

Make's product roadmap has increasingly focused on AI capabilities, including the next generation of AI Agents and transparency features that show how agents make decisions. The Miriwa results provide a case study for how automation platforms like Make, Zapier, and n8n can be applied internally before being sold to customers.

For organizations evaluating AI adoption, the Miriwa program offers a template: start with internal use cases, measure outcomes, and build the surrounding processes before scaling. Make's approach prioritizes structured work, clear success metrics, and governance over broad deployment, which aligns with how most companies are navigating AI implementation in 2026.