In Windows, searching for files is often like a lottery – even an SSD does not save when the system spends hours scanning terabytes of data. Lertaro solves this in a radical way: instead of going through folders stupidly, the program pulls data directly from NTFS MFT and USN Journal, indexing changes in real time — and returns results in milliseconds, even if you search for “report_2026_draft_final_v2.docx” among thousands of similar titles. And there is also a fuzzy search with exceptions, a preview of files through QuickLook and the ability to work on top of any window — if you need to quickly find a document during a video conference.
What is NTFS MFT and why it speeds up file searches
The NTFS Master File Table (MFT) is the heart of the NTFS file system. Think of it as a giant table, where each row is a file or folder, and the columns contain all the metadata: name, size, date modified, access rights, and even the physical location on the disk. Unlike old file systems, where folders had to be traversed recursively, MFT allows you to get a list of files in a matter of milliseconds — just read one file on the disk.
Programs like Lertaro use MFT in conjunction with the USN Journal, an NTFS change log that records every file operation (create, rename, delete). Thanks to this, the index is updated instantly, without slow crawling. For example, if you’ve renamed 10,000 files, Lertaro will find out in seconds, not minutes like a standard Windows search.
The advantages are obvious: fuzzy search (“dok” will find “document.pdf”), pattern exclusion (“.tmp”), integration into Explorer or file selection dialogs, plugins for extensions. And all this — without a background process that eats up resources. MFT isn’t magic, it’s just making good use of what’s already in the system.
How Lertaro improves file experience: features and integration
Lertaro is not just a launcher, but a tool that turns working with files on Windows into instant response. Thanks to NTFS MFT and USN Journal analysis, the program finds files in milliseconds, bypassing the slow recursive folder search. Even in repositories with hundreds of thousands of files, the result appears before you can take your finger off the keyboard.
Fuzzy search works smart here: you can search by parts of the name (“prj rpt 2026” will find “Project_Report_Q2_2026.xlsx”), exclude words with a minus (“-temp”) or require exact matches in quotes. Errors in the query? The system will suggest the nearest options, as search giants do, but locally — without sending data somewhere in the cloud.
The QuickLook preview opens directly in the Lertaro interface: hover the cursor over the file — and you’ll see images, text content, or even the first frames of a video. No need to launch programs or wait for Explorer to load. And thanks to the “on top of windows” mode, the search will not close even if you switch between applications.
Integration with the system is deep: you can call Lertaro from the Explorer context menu, replace the standard file selection dialog in any program, or bind hotkeys to specific actions. For example, Ctrl+Shift+F will open a search in the current folder, and Ctrl+Enter will launch the file immediately in the associated program.
The functionality is expanded with plugins: from cloud storage indexing to specialized previews for CAD files or archives. For example, if you need to quickly find shell commands, you can use integration with a tool like Navi, which we wrote about earlier. Developers have already created modules for working with Git, searching the clipboard history, and executing PowerShell commands directly from the results. Written in C# using .NET 10 and WPF, the app runs fast even on low-end machines, and the index updates in real-time—change a file and it’s instantly searchable.
Technical features: C#, .NET 10 and real-time indexing
Lertaro is written in C# under .NET 10 — this gives access to the latest runtime optimizations, including AOT compilation and SIMD instructions for data processing. WPF is not just for beauty here: thanks to DirectX rendering, the interface does not slow down even during mass updating of the result lists. The asynchronous architecture allows you to keep the UI responsive while background threads are churning out MFT or USN Journal.
Real-time indexing is not marketing. The program picks up file changes via the USN Journal (updates every 50-200ms depending on disk load) and instantly updates the internal cache without a full reindex. For comparison: traditional launchers like Everything do a full scan of the disk at the first launch – this is not the case here. Fuzzy search works on a modified Levenshtein algorithm with caching of intermediate results, which allows you to find “dcmnt.txt” for the query “dokument” even in 500 thousand files in <100 ms.
Optimizations are not limited to search. The QuickLook preview is loaded in parallel with the creation of the list of results, and plugins (for example, for integration with Git or 1Password) are connected via MEF – this allows you to extend the functionality without recompiling the main code. Even working on top of other windows is implemented through low-level Win32 API hooks to avoid delays when switching focus.
- NTFS MFT + USN Journal: indexing without full disk scan, real-time updates.
- .NET 10 + AOT: fast start (cold start <300 ms) and low memory consumption (~50 MB for 1 million files).
- WPF + DirectX: smooth UI even with 100+ results in the list with preview.
Why local MFT search is important for AI and SEO professionals
SEOs analyze gigabytes of logs, CSV files with keywords or reports from tools like Ahrefs every day. When searching for a file takes 10 seconds instead of a minute, savings are measured in hours over thousands of requests. And if you need to find a specific log by date or keyword among hundreds of folders? Here, an MFT search like Lertaro gives results instantly – even if the file name is misspelled or partially.
For AI professionals, quick access to datasets is not just a convenience, but a critical factor. Imagine you’re looking for a data preprocessing script written six months ago, but you only remember that it had “augment” and “2025” in the title. Instead of digging through dozens of folders, fuzzy search returns results in a second. And if you need to check the contents of the file without opening it — the QuickLook preview will show the first 50 lines of the dataset or the Python code right in the search window.
Integration with File Explorer or file selection dialogs means the new tool doesn’t disrupt your usual workflow. Plugins allow you to expand the search to cloud storage or databases — for example, to search for scripts simultaneously in local folders and on GitHub. Indexing time? On an SSD with 500K files, less than a minute, and real-time updates track changes via the USN Journal. For those who work with data, this is not just an acceleration – it is a change in the way they work.
Andrey Krasovskiy is a programmer and data scientist experienced in building complex automated systems with Python, Google Colab and n8n. His expertise spans SEO ecosystems, API integrations (Ahrefs, Google Ads, Search Console) and content pipelines. Andrey combines technical precision with an entrepreneurial mindset to build solutions that deliver real results.