Ramp created an internal AI tool that acts as a wrapper around an LLM. It's connected to Notion, Slack, and Snowflake, building a persistent memory of team activities and individual work styles. This "company brain" can diagnose business issues, summarize communications, and draft meeting prep in minutes, not weeks.

For Wellington Management, the most critical AI use case is not a generic tool but a system to mine its proprietary IP: a data lake of notes from over 20,000 annual company meetings. This turns decades of institutional knowledge into an interactive, queryable asset for its investors.

By the end of 2026, recording every meeting and applying AI agents to transcribe, summarize, assign action items, and align with strategy will be table stakes. Hoffman argues that companies not doing this will be making excuses, akin to sticking with horse-drawn carriages in the age of the car.

Remote work's inherent documentation—recorded meetings and transcripts—creates a comprehensive dataset ideal for training a corporate AI 'brain.' In contrast, in-person work loses valuable context from unrecorded hallway conversations, leading some founders to re-evaluate their return-to-office mandates.

A key practice at OpenLoop is making all meeting transcripts available to everyone in the organization. This radical transparency creates a massive, shared knowledge base that can be queried by AI systems, allowing employees to access information and context from meetings they didn't attend.

Instead of manual note-taking, use AI tools to transcribe and summarize all meetings. This creates a unique, searchable knowledge base from your conversations, which can be leveraged to improve preparation, follow-ups, and decision-making over time.

Within three years, the default for all enterprise meetings will shift to "record on." This ambient data capture will feed a new system of intelligence, automatically extracting insights, monitoring for compliance risks, and diffusing issues proactively. Unstructured conversation data will become a core enterprise asset.

Future AI models will learn complex, multi-step tasks by watching screen recordings. Companies should begin capturing video of their key internal workflows now. This data, which is currently discarded, will become a valuable proprietary asset for training AI agents to automate bespoke business processes.

Firms that meticulously document the reasoning behind trading decisions are building a proprietary dataset for future AI agents. This intellectual property, capturing the firm's unique philosophy, will be invaluable for training AI that can truly understand and operate within its specific context, forming a powerful competitive advantage.