FAQs | Enfacta

Frequently asked questions

Learn how Enfacta uses AI to improve your workflow, and how it differs from cloud AI.

Does my information ever leave my computer?

▫ No. Enfacta runs entirely on your device.
▫ Your documents, queries, stored facts, and workspace data are never uploaded or sent to external AI services.
▫ By contrast, cloud-based AI tools require you to send your data to remote servers for processing. This introduces additional security, privacy, and compliance considerations that are outside your direct control.
▫ Enfacta avoids cloud-related security and compliance issues entirely by keeping all processing local on your computer.

How does Enfacta use AI without the cloud?

▫ Enfacta runs modern AI models locally on your computer.
▫ When you ask a question, Enfacta retrieves the most relevant sections of your files using retrieval logic engineered for real-world professional documents, and reasons directly over that evidence.
▫ Cloud-based AI tools typically rely on sending large amounts of context — documents, conversation history, and prompts — back and forth over the network.
▫ Enfacta eliminates the need to send large amounts of information to and from the cloud by bringing the AI to your data.

Can I use Enfacta like a normal chatbot? How smart is it?

▫ Yes — Enfacta accesses modern models' extensive general knowledge and can be used for general questions and discussion.
▫ Where Enfacta differs is how it retrieves and grounds information.
▫ Cloud-based chatbots rely on accumulated conversation context. This can cause responses to drift over time.
▫ Enfacta prioritises direct, factual answers drawn from your documents or stored facts. Each question triggers a fresh, targeted retrieval pass — keeping answers stable, inspectable, and grounded in trusted material.

What is a Workspace, and why does it matter?

▫ Yes — Enfacta accesses modern models' extensive general knowledge and can be used for general questions and discussion.
▫ Where Enfacta differs is how it retrieves and grounds information.
▫ Cloud-based chatbots rely on accumulated conversation context. This can cause responses to drift over time.
▫ Enfacta prioritises direct, factual answers drawn from your documents or stored facts. Each question triggers a fresh, targeted retrieval pass — keeping answers stable, inspectable, and grounded in trusted material.

Can I tell Enfacta what information is authoritative?

▫ Yes. Enfacta allows you to store explicit facts within each Workspace.
▫ These facts are treated as authoritative. They are prioritized over documents, conversational context and general knowledge.
▫ Most cloud-based AI tools do not offer a way to formally distinguish between authoritative truth and conversational input. This can lead to inconsistencies and drift.
▫ Enfacta gives you a way to anchor the system to what you know is correct.

How does Enfacta decide what information to use when answering?

▫ Enfacta follows a defined hierarchy:
ㅤㅤ1) User-stored facts (always authoritative)
ㅤㅤ2) Relevant document excerpts
ㅤㅤ3) General model knowledge
▫ Document retrieval is purpose-built to reflect how professional documents are written.
▫ In many cloud-based systems, these sources are blended together implicitly, making it difficult to understand why an answer was produced.
▫ By applying this heirarchy, Enfacta's answers are consistent, inspectable, and grounded in identifiable sources.

What types of documents can Enfacta work with?

▫ Enfacta is designed to accept and reason over a wide range of common documents:
PDF, Word, plain text, Excel, CSV, and images that contain text.
▫ Document text is extracted, structured, and indexed locally so all supported files can be searched and queried within a Workspace.
▫ For spreadsheets, Enfacta understands tabular data and can answer questions and generate graphs about totals, trends, and comparisons.
▫ For images, text is detected and converted into readable material by OCR (optical character recognition) that can be indexed and queried like any other document.

Can Enfacta use our entire company knowledge base as context? It's huge.

▫ Yes. Enfacta uses a retrieval-based approach designed for large, real-world knowledge bases.
▫ Your documents are indexed and broken into meaningful sections, preserving structure rather than flattening into generic text .
▫ When you ask a question, Enfacta searches the entire knowledge base and retrieves the information that best matches the intent of the question.
▫ The AI reasons over this specific evidence, producing answers that are accurate, grounded, and complete — without overwhelming the model.

How does Enfacta perform compared to cloud-based AI?

▫ Because it runs locally, Enfacta avoids bottlenecks common to cloud-based AI tools such as network latency, shared server queues, and accumulating context.
▫ Enfacta performs targeted retrieval for each question, keeping performance and answer quality consistent over time.
▫ Enfacta keeps a per-session rolling context window to keep the memory fresh and not overloaded.
▫ Modern 8GB+ laptops typically handle this workload well. If your computer comfortably runs professional software, Enfacta will feel responsive. GPU- and Apple Silicon-accelerated models are offered when that hardware is detected.