"Modern management is completely broken," says Alex Kvamme, CEO of Pathlight. "Today's managers are expected to be data analysts as well as career mentors, but they lack the skills for the former and the time for the latter. And the problem isn't getting any easier. Teams are getting bigger as companies feel pressure to streamline, and the constant threat of layoffs is driving employee anxiety to new highs. The only way forward is to overhaul the role of the modern manager and redefine it for the 21^st century."

Kvamme continues, "By harnessing the power of large language models, AI Manager can handle the analytical side of management, enabling human managers to focus on their most important functions: leading and supporting their teams. The result is that employees get the best of both worlds: an always-available, always-objective AI Manager to help them understand how they are doing today and a human manager with the time and experience to help them achieve their career goals tomorrow."

Powered by the latest advancements in neural networks and Pathlight's performance-data infrastructure, AI Manager focuses on the two major types of performance management for customer-facing teams: reviewing and analyzing customer conversations and analyzing and course-correcting performance metrics.

Accessed via a chat window anywhere in the Pathlight platform, AI Manager can proactively alert the user when it has detected a customer conversation or KPI trend that is worth addressing. AI Manager is also available around the clock to answer questions that an employee might have about, e.g. recent performance or customer conversations.

A significant obstacle for businesses when creating AI-driven applications is training the underlying LLM with enough relevant, company-specific information to work effectively. This is made more difficult by the model's inherent context window and token limitations, which prevents enterprises from simply uploading the thousands of customer conversations and metric data points that are created every day.

"Pathlight uses sophisticated methods that enable AI Manager to efficiently access the right data, while staying within processing limits and keeping customer data secure," says Trey Doig, CTO of Pathlight. "This approach allows AI Manager to consistently improve its performance by learning from examples of successful and unsuccessful behaviors. This innovative solution not only enhances the AI's output, but also helps it adapt and fine-tune its recommendations based on a company's operations and employee performance."

The first version of AI Manager will be available in a closed beta and is designed for the frontline employee to use. In the coming weeks, Pathlight will make AI Manager available to managers who want to streamline and automate their time-consuming workflows.