Keepi for Web Apps

Keepi, created by Vasilije Markovic, is a web-based personal knowledge assistant for capturing, indexing, and retrieving personal content. The app collects links, notes, images, and audio then makes that material searchable through an AI query interface. Its core workflows emphasize low-friction capture and automatic indexing, helping researchers, knowledge workers, students, and solo professionals turn scattered digital items into an organised, searchable memory.

What tasks can you actually use it for?

Keepi targets personal knowledge management tasks: saving web URLs, forwardable messages, images, documents, and voice notes into a single archive. The app accepts content via messaging and a browser-based interface, then indexes items so they can be retrieved later. Typical use cases include collecting research links, archiving meeting snippets, and keeping reference material that you can query instead of leaving as passive bookmarks.

How reliable are the summaries and transcriptions?

Summaries and transcribed text are the app's primary outputs. The tool produces article summaries and converts forwarded voice messages into searchable text. User feedback highlights effective handling of clear voice notes, while transcription accuracy decreases with heavy background noise or strong accents. For long-form articles, the summaries extract key points but should be verified against the source when precision is required.

What input formats and workflow constraints matter?

The app accepts a broad set of multimedia inputs. Supported content types include web URLs, text snippets, images, documents, and voice messages. Capture is done through forwarding or the web interface; incoming items are automatically indexed and categorised. That WhatsApp-first capture model suits people who already use messaging for quick saves, while the browser access covers desktop clipping and manual uploads.

How does it treat privacy and how does it compare to other tools?

Operationally it runs as a web service integrated with messaging. Because the app processes uploads and messages through its web layer and messaging channel, users handling sensitive material should review the developer's data policies. The developer's background includes data engineering and machine learning experience and an open-source project focused on knowledge management, which explains the product's focus on personal archives rather than general conversation.

A clear fit for message-driven knowledge capture, with cautious verification advised

The app is a practical option for people who prioritize quick, message-based capture and searchable archives for research or personal projects. Expect useful but imperfect AI outputs, so verify transcriptions and summaries before using them in formal work. For sensitive content, review data-handling details; the app suits users who accept a web-based capture flow and want a conversational way to retrieve their saved material.