Frequently Asked Questions

Epimonos is a Personal AI Workspace that helps individuals and teams organize memory, context, meetings, documents and knowledge across multiple AI systems.

Instead of focusing on a single AI model, Epimonos focuses on continuity: keeping your work connected over time.

Most AI tools are strong at answering isolated questions but weak at maintaining long-term context across projects, meetings and documents.

This leads to fragmented workflows where knowledge is scattered across tools. Epimonos is designed to solve this by creating a persistent layer of memory and context.

The Personal AI Workspace is a unified environment where your AI tools, documents, conversations, meetings and knowledge are connected.

It replaces tool-switching with a single continuous workspace.

A few examples from practice, close to the idea of a personal assistant:

• An automatic daily overview of open action items, important emails and your calendar
• Summarizing and searching large documents or export files, without the content ever leaving your organization
• Having contracts and terms reviewed for notice periods, privacy clauses and cost
• Keeping track of open quotes and invoices, so nothing falls through the cracks
• Preparing administrative work — for example processing incoming receipts and drafting invoices in advance, sent only after your approval
• Flagging routine checks, such as whether systems are up to date and reachable

Epimonos adapts to your workflow, not the other way around.

Epimonos is a good fit if AI is part of your daily workflow and not just occasional use. You benefit most if you want:

• Continuity across projects over weeks or months
• Structured memory instead of chat history
• Meeting capture and synthesis
• Shared context across teams
• Multi-model flexibility without fragmentation

Typical users are operators, consultants, founders and knowledge workers who deal with ongoing complexity rather than single tasks.

Epimonos is not designed for people whose primary interest is building or tweaking AI infrastructure. If your goal is to experiment with models, test setups or optimize inference performance, you will likely prefer a more hands-on or DIY stack.

In that case you probably enjoy: running local models and benchmarks, fine-tuning or switching models frequently, building homelab AI systems, or managing GPUs, servers and containers directly.

That is a valid direction, but it solves a different problem than Epimonos.

Some AI enthusiasts will appreciate Epimonos, especially when working across multiple models. However, Epimonos is not primarily an experimentation platform — it is a production workspace for knowledge and execution.

A few examples from practice:

• Lawyers — forwarding a prompt to a US-based frontier model can lead a court to treat that exchange as sharing information with an external party, affecting attorney-client privilege.
• Accountants — offering a client's books to a US AI service for analysis shares commercially sensitive information outside the EU, discoverable under the US Cloud Act.
• Consultants for government or critical sectors — pasting project documents into a US model can breach procurement or confidentiality terms that require EU-only processing, and may be relevant under NIS2.
• Healthcare providers — analyzing patient data with a US AI service processes special-category personal data (GDPR Art. 9) outside the EU, and cuts against medical confidentiality.
• Cybersecurity consultants — sharing incident or vulnerability data with an external AI platform adds another party to the already-sensitive chain around a live incident or pentest, often in breach of the governing NDA.
• Companies with export-sensitive or strategic data — sectors like defense, semiconductors and pharma often have explicit policies that sensitive information must not leave the EU, sometimes imposed by their own customers or regulators.

Epimonos is designed to remove that trade-off: sensitive documents and conversations stay inside your own, encrypted environment. See our security page for how that works technically.

No. Different AI solutions solve different problems. The right choice depends on what you are trying to achieve.

Some tools are optimized for conversation, others for infrastructure, others for local execution, and others for long-term knowledge management. Epimonos focuses on the layer above all of that: continuity.

ChatGPT and similar assistants are excellent for fast answers, writing and general assistance. However, they are not designed to maintain deep, persistent context across projects, teams and long-term workflows.

And you don't have to choose: Epimonos combines using ChatGPT and Claude with a local model. Confidential documents and personal data are processed locally; only non-identifiable, aggregated results can — if you choose — still be shared with the commercial models.

Claude is a strong general-purpose AI assistant with excellent reasoning and writing capability. Like other standalone AI assistants, it does not provide a unified long-term workspace for memory, meetings, documents and multi-project continuity.

Same as with ChatGPT: you don't have to choose. Epimonos lets you combine Claude with a local model, so confidential documents stay local and only non-identifiable, aggregated results are shared if you want.

OpenWebUI is a powerful interface for running local models and connecting to different backends. It is primarily focused on model access rather than structured knowledge, workflows and long-term context management.

AnythingLLM is a strong tool for document-based AI and retrieval workflows. It is effective for working with knowledge bases but is not primarily designed as a full cross-project workspace for teams, meetings and persistent contextual memory across systems.

LibreChat provides a flexible chat interface for multiple AI providers. It focuses on chat orchestration rather than structured memory, project continuity and workspace-level intelligence.

An AI PC or device such as RTX-based systems can run models locally and offer full control over inference. This is a strong option for users who want offline execution and full ownership of their hardware stack.

However, it does not inherently solve higher-level problems such as: long-term memory across projects, team collaboration, structured knowledge management, meeting intelligence, or cross-model orchestration.

Epimonos focuses on these higher-level layers rather than raw compute.

Running local models provides privacy, control and independence from cloud providers. However, local models alone do not provide structure around memory, workflows, collaboration or persistent context across tools and time.

Yes. Local AI hardware can be used as part of the Epimonos ecosystem. The long-term vision is to support multiple inference backends, including cloud-based models, local LLM runtimes (Ollama-style setups), and AI workstations and appliances.

Epimonos focuses on orchestration, memory and context rather than the model layer itself.

There is no universal best choice. Each category solves a different problem. The most important distinction is whether your primary goal is using AI as a tool for work and knowledge, or building and managing AI infrastructure.

None, if you use the powerful local models that come standard with your workspace. You can also easily connect your existing ChatGPT or Claude subscription.

Do you need premium frontier models for more complex questions? You can unlock those via an API — you choose: connect your own API account, or use our central API connection. In the latter case, those costs are billed separately, or you can purchase credits (tokens) in advance.

Epimonos is designed to support flexible deployment models, including cloud-based and hybrid configurations.

Yes. Everything you put into Epimonos lives on a separate, encrypted volume (LUKS2, AES-256-XTS). Unlocking is tied to your own machine — the volume won't open anywhere else. You receive a recovery key once at handover; we don't keep it ourselves.

See our security page for the full picture, including what we can and can't see.

Yes. Epimonos is designed to support Bring Your Own Key (BYOK) setups, allowing direct connections to model providers.

Local or self-hosted deployments are part of the long-term roadmap. The goal is to allow flexibility between local, hybrid and cloud setups.

The architecture is designed to minimize vendor lock-in through multi-provider support, BYOK integration, support for local models, and portable data concepts.

The goal is that your data and knowledge remain accessible and portable. Epimonos is designed to earn long-term usage rather than enforce dependency.

The Personal AI Cloud is a future deployment model where users can run their own private AI workspace while leveraging shared or external AI compute resources. It combines three principles: data ownership, deployment flexibility, and shared AI infrastructure efficiency.

The goal is to give users control over their environment without requiring them to manage low-level infrastructure complexity.

The long-term vision of Epimonos is to become the layer that connects AI models, knowledge, memory, documents and workflows into one continuous workspace. Not as a model provider, but as a coordination and context layer across models, tools and environments.

One workspace. All your AI. Your memory and context.

New AI models appear every week, and the ecosystem evolves rapidly. There is no shortage of high-quality models, and that number will only increase.

We are not trying to build the next model. We are building the layer that connects the most useful models together with your documents, memory, structure and workflows. Models will change. Your context should not depend on them.

If your main goal is experimenting with AI models, infrastructure or hardware, a DIY or local-first approach is likely a better fit.

If your main goal is using AI to improve how you work across projects, teams, meetings and knowledge, Epimonos may be a better fit.

Most AI tools focus on one of three layers: models, interfaces, or infrastructure.

Epimonos focuses on a fourth layer: continuity across models, tools, teams and time. It is not designed to replace AI models or infrastructure. It is designed to make them work together.