Glean – Weekly Recap

Glean is an enterprise AI search and knowledge management platform, and this weekly recap reviews the company’s latest leadership moves, product updates, and ecosystem activities. The week underscored Glean’s focus on scaling its AI agent framework, improving cost and latency, and preparing for a phase of what it calls hyper growth.

Glean appointed veteran technology executive Amar Maletira as Chief Operating Officer, emphasizing his prior roles as CEO and President of Rackspace Technology and EVP & CFO at VIAVI Solutions. The hire signals a push toward stronger operational discipline, customer focus, and scalable processes as Glean targets larger enterprise deployments.

On the product side, Glean introduced Waldo, its first specialized agentic search model built on NVIDIA’s Nemotron 3 Nano and post-trained for search planning. Waldo is designed to orchestrate multi-step enterprise queries, select tools, and decide when to rely on larger reasoning models, targeting retrieval-heavy workloads where latency and cost are critical.

Internal benchmarks cited by the company indicate Waldo delivers roughly 10x faster per-call performance, with about 250 ms P50 latency versus roughly 3 seconds for Glean’s default reasoning model. In Glean’s own harness this translates to about 50% lower end-to-end latency and roughly 25% lower token costs, suggesting improved unit economics and scalability for high-volume customers.

Glean also highlighted a third-generation AI harness architecture that treats the harness as a distributed context management system for complex, longer-running workflows. By progressively loading instructions into skills rather than relying on a large static system prompt, Glean reports cutting its system prompt size by more than 45%, aiming to improve accuracy and task completion.

These infrastructure-level advances position the platform to handle more tools, searches, and executions without overwhelming model context windows. For enterprises, the changes are intended to support more reliable production-grade agents, potentially lowering compute costs while enabling more sophisticated use cases.

Customer-side adoption was showcased through an Intercom use case, where a senior content designer built a custom VERBI agent on Glean’s platform. Trained on Intercom’s full content design system, VERBI centralizes style, tone, and copy guidelines, giving product designers immediate access to consistent content standards.

This example underscores Glean’s ability to support role-specific agents that embed directly into design and product workflows. Such deployments may enhance product stickiness, support upsell opportunities, and strengthen Glean’s positioning in knowledge-heavy functions seeking to streamline documentation and guidance.

The company also promoted its upcoming Glean:LIVE event on May 12, focused on moving from scattered AI experiments to agents delivering measurable business impact. Speakers from Glean and HubSpot will discuss frameworks for tying AI agents to quantifiable outcomes, reflecting Glean’s emphasis on ROI in budget-conscious enterprise environments.

Across leadership, product, and go-to-market activity, the week illustrated Glean’s push to pair seasoned operational management with deep technical innovation in AI agents and search orchestration. Collectively, these developments suggest the company is working to solidify its infrastructure, expand enterprise relevance, and support more durable, efficiency-driven growth.