Coworker.ai launches OM2: the organizational Memory Layer that makes Enterprise AI 9x cheaper. Already trusted by 300+ companies, Coworker.ai delivers organizational memory that makes enterprise AI 9x cheaper, 64% faster, and with answers preferred on quality 84.5% of the time. SAN FRANCISCO, CA - WEBWIRE - Thursday, September 3, 2026 Coworker.ai today announced the launch of OM2, an organizational memory layer designed to reduce enterprise token spend. Instead of rebuilding company context from scratch every session, where over 50% of enterprise token budgets are spent on context retrieval alone, OM2 learns continuously across enterprise data. The result is a permission-aware, always current understanding of a company that every AI model, human, and agent can draw on. Because the layer is precomputed rather than reconstructed every session, OM2's context layer alone drives a 9x reduction in token spend, before any additional savings from model routing. Enterprise AI Problem: AI with no memory and model lock-in Every AI your company uses today starts from zero: every session, every tool, every time. Agents re-discover the same context, burn tokens relearning what the business already knew, and guess at which tool to call next. Further, companies want sovereignty over their data and the ability to take that context with them, not lock in to a single vendor's ecosystem. Coworker.ai's OM2 Solution: The Self-Building Memory Layer for Enterprise AI OM2 continuously ingests data across 50+ enterprise platforms, building what the company calls a neural graph: a dense, permission-aware memory that is orders of magnitude denser than traditional document-level knowledge graphs. As new information arrives, whether a Slack thread, a Salesforce update, or a meeting note, OM2 extracts it into discrete, linked facts identifying who was involved, what was decided, and which project or account it touches. Those facts are computed once and reused across every query, rather than re-derived by every model, every session. Serving as a universal intelligence layer, OM2 gives every human, agent, and AI application one unified, permission-aware picture of your business. Coworker.ai customers like RapidSOS are already using OM2 today. "We are huge proponents of Coworker. Before Coworker and OM2, keeping up with how fast the business moves meant our RevOps team spent hours stitching together Salesforce, Slack, and meeting notes to trust an answer," said Anna Waring, Director, Revenue Operations and Systems. "Now the answer is already there, current and correct, the moment we ask." Key Features of OM2 include: * Move faster, spend less: OM2's context layer alone drives a 9x reduction in token spend, by computing and reusing facts instead of re-deriving context every session. * Intelligent model routing: Layer in Coworker's Optimized Routing, which sends each task to the most efficient model, and cost savings reach 51x. Instead of brute-force tool calling, OM2 lets Coworker route intelligently across models. * Answers your AI can't reach today: OM2 stores unstructured information as structured facts, with answers preferred on quality 84.5% of the time. * Works with the AI you already use: OM2 connects to Claude, ChatGPT, Gemini, Perplexity, and custom agents out of the box so organizational memory is never lost. * Security in every fact: Access policies are built into every fact and connection, on isolated single-tenant infrastructure (SOC 2, GDPR, CASA Tier 2) so everyone sees only what they're allowed to see. * One memory for every team: Every person and agent draws on the same shared, living picture of the company, so decisions stop getting lost between tools and teams. "Token costs are exploding, and most AI still doesn't understand how to be effective inside a real company," said Alex Calder, Co-Founder and CEO of Coworker.ai. "OM2 is continuously learning and understanding your business, while still enforcing strict permissions at every node. That's how we get to 9x cheaper before routing even enters the picture, we cut the cost of relearning your business from scratch, not the corners. And unlike a lot of the AI industry, we're not incentivized to keep you burning more tokens. We're incentivized to help you use fewer." Try Coworker.ai's OM2 today at https://coworker.ai/ About Coworker.ai Coworker.ai is the enterprise AI platform built on its proprietary organizational memory layer, OM2. OM2 understands your enterprise and gets smarter with every interaction, driving 9x cheaper AI. Coworker.ai also routes every request to the right model, and is available headlessly through its MCP or via native apps. Coworker.ai is 64% faster and preferred over standard tool-calling approaches 84.5% of the time. Founded by former Uber executives Alex Calder and Bradford Church, Coworker.ai is backed by leading Silicon Valley investors, including Jeff Huber, former SVP of Google, and Ramtin Naimi. Learn more at https://coworker.ai/ WebWireID359968 * Contact Information - Riley Munks - PR Advisor - Activate PR - https://www.activate-pr.com/ - Contact via E-mail This news content may be integrated into any legitimate news gathering and publishing effort. Linking is permitted. Distribute your news. * Every day, hundreds of individuals and companies choose WebWire to distribute their news. * WebWire places your news within numerous highly trafficked news search engines generating leads and publicity. * Submit Your Release Now!

Ex-Uber managers Alex Calder and Bradford Church raised $13 million in seed funding for their startup, Coworker.ai, which develops AI teammates for workplace tasks. The funding round was led by Jeff Huber of Triatomic Capital, with participation from Abstract Ventures, Operator Collective, Eniac Ventures, and K2 Access Fund. Coworker's AI agents are designed to handle tasks across various functions, freeing human employees for more creative work. The startup previously raised $3.5 million in pre-seed funding.