Online meetings are increasingly joined by bots that automatically take notes and observe from the cloud. But is there a better option?

Quill Notes Inc., or Quill Meetings, the maker of the Quill app and Quilliam, thinks it has an answer: a local-first “chief of AI staff” that lives on-device, takes notes, coordinates context and tools, and learns how you work. Nothing you do leaves your device without your permission.

Quill co-founder and Chief Executive Michael Daugherty told SiliconANGLE in an interview that Quilliam got dubbed “Chief of AI Staff” because modern-day workplaces have become a zoo of AI agents and tools.

“It’s gone from me typing code to me just this morning trying to manage five different cloud code agents that are doing different things,” he said. Not only does Quilliam act as a note-taker, but it can also connect to other tools, configure itself and send them information after ingesting and understanding those notes.

Unlike other AI tools that lock users into a narrow workflow on cloud infrastructure, Quill aims for configurability and control. Users can choose where data lives, either locally or backed up to the cloud; how inference runs, either on their local device, on-premises for sovereignty or in the cloud for a smarter agent; and configure how their workflow evolves over time.

The company recently raised a $6.5 million seed round led by Basis Set Ventures, with participation from 500 Global, Naval Ravikant, Morado Ventures and AME Cloud Ventures. It also launched its app for macOS and Windows, with an iOS companion app available to paid users.

At the core of Quill is data control. User data and, when possible, compute live on the user’s device and Quill can operate without sending anything off the machine — or off the network — unless the user chooses otherwise.

Recorded audio never leaves the device. Transcription and insights can run locally; cloud sync is optional and encrypted when enabled. Users can also choose where AI inference happens. Even when a request is routed to a cloud model, the company says the data is not used to train models.

“We do the transcription on your device… your transcripts live there. The audio never even leaves,” added Daugherty.

On device, broad audience

Quill is aiming for a broad audience: prosumers who want an agent that actually does work across their tools and regulated, privacy-sensitive organizations that can’t afford to pipe meeting data into the cloud. The strongest pull, Daugherty said, is from customers focused on data sovereignty, especially in Europe, where compliance expectations make “where the data lives” as important as what the AI can do.

As an agentic AI helper, Quilliam is designed to be proactive and connect with other tools via Anthropic’s Model Context Protocol, a connector that can hook into third-party services. That allows Quilliam to act less like a passive notetaker and more like a teammate that moves information through a user’s daily workflow — across tools such as Notion, Linear, Affinity, Obsidian, Airtable and Gamma.

For example, after a product meeting, Quilliam follows a workflow where it listens to the conversation, transcribes and processes the notes, analyzes the discussed action items and then automatically creates or modifies Linear tickets, updates Notion documentation and drafts updates in Airtable. Before a client call, it reviews past interactions, assembles briefing materials and surfaces relevant context. Over time, it observes user preferences, builds tailored automations and even suggests improvements based on learned patterns.

Quill’s pitch is that this becomes a living system: evolving, learning and functional, offering its own recommendations for customizing interface, integrations and routines.

With the funding last week, Quill is putting most of its energy into expanding its team and hardening the product for the next wave of free users. Daugherty said the company is hiring, especially around React Native, to make the mobile app cross-platform and more robust while focusing on enterprise requirements for regulated environments.

The platform includes a free app that works locally without a subscription and still offers a powerful baseline. To access cloud storage and higher-end inference options, users can subscribe to Pro tiers that add backup and expanded capabilities.

Subscriptions start at $6.99 per month, billed annually, for Lite and $16.99 per month, billed annually, for Unlimited.

Images: SiliconANGLE/Microsoft Designer, Quill Notes