Windows Early Access

Find answers across private documents you add to MemoRay — without sending them to cloud AI tools.

The first MemoRay early-access release is a Windows app that turns PDFs, scans, receipts, IDs, contracts, and other private records into a searchable local memory on your device, with answers grounded in your own files.

Join the Waitlist See how it works

Windows early access stays limited while onboarding remains hands-on.

Founder Access onboarding is limited and hands-on.

Find records faster Ask across your files Verify sources No document upload for core workflow

Why MemoRay

Stop digging through folders, filenames, and scattered records.

Find records faster

Search receipts, invoices, agreements, IDs, statements, and scans using natural language.

Ask across your files

Get answers grounded in imported documents instead of copying sensitive content into generic chat tools.

Verify where answers came from

MemoRay is designed around evidence-backed results and visible sources.

Early Access Preview

A look at the current Windows early-access app.

ask memoray local retrieval early access preview

Ask MemoRay is retrieval-focused and grounded in imported files.

OCR can help with scans and images; result quality still depends on source quality.

Availability

The first early-access cohort is opening on Windows.

Join the waitlist for Founder Access and early onboarding.

Real Use Cases

Search the records you already have — without handing them to cloud AI.

A lot of useful personal and small-business information sits in invoices, agreements, receipts, tax records, policies, warranties, IDs, statements, scans, screenshots, and notes. Those are also the records people hesitate to paste into cloud AI tools because they do not fully trust where the data goes. MemoRay lets you import the files you choose, then search or ask questions with source-visible results — built around local processing, evidence-backed answers, and more user control.

Find tax records faster

Search tax PDFs, property tax statements, 1099s, W-2s, receipts, and deduction-related records without digging through folders.

Example queryFind my 2023 property tax statement.

Retrieve receipts by merchant or date

Find receipts across PDFs, screenshots, scans, and images using natural language instead of exact filenames.

Example queryShow my Panera receipts from January 2026.

Bring up IDs and personal records

Find passport scans, driver's license copies, forms, letters, and other sensitive personal records you prefer not to upload.

Example queryFind my passport renewal scan.

Search invoices and warranties

Locate invoices, purchase records, warranties, and service documents when you need proof of purchase or support history.

Example queryFind the washing machine invoice for the warranty claim.

Find signed agreements

For consultants, founders, and micro-business owners, retrieve signed agreements, proposals, statements of work, and client records.

Example queryBring up the signed client agreement for Acme.

Ask about policy or coverage details

Ask grounded questions across insurance, benefits, financial, or policy documents and verify the source behind the answer.

Example queryWhat deductible did I choose on my auto insurance?

These examples are retrieval scenarios, not a promise that every file will produce a perfect answer. Result quality depends on file quality, OCR quality, and whether the answer is present in the imported documents.

Join the Waitlist See how it works

Trust Foundation

Local core workflow

Import, OCR/extraction, indexing, search, and retrieval are designed to run on your device for the core local workflow.

No document upload required

Core MemoRay functionality does not require uploading document content to a cloud service.

Evidence before fluency

MemoRay prioritizes grounded, source-visible results over unsupported AI responses.

Core Capabilities

Import documents, search locally, and ask grounded questions.

Import

Bring in PDFs, DOCX files, and images.

Supported import types include PDF, DOCX, PNG, JPG, and JPEG files.

Search across records with local retrieval.

Use search and filters across imported files without manually digging through folders.

Ask

Ask questions grounded in your files.

Ask MemoRay returns evidence-backed results tied to imported documents.

How It Works

From import to answer, MemoRay keeps the workflow local and visible.

Import documents, process them locally, then search or ask questions with sourced results.

File Selection To Retrieval

A transparent local workflow from file selection to retrieval.

MemoRay turns selected local files into searchable, evidence-backed results.

Select

Select files or folders and review staged documents before ingest.

Ingest

OCR, extraction, chunking, indexing, and duplicate detection run locally.

Search

Use hybrid retrieval with lexical matching, local vectors, and filters.

Ask

Ask questions and get results grounded in imported files.

Protect

Use STANDARD and VAULT visibility controls for sensitive records.

PDF, DOCX, and image support Local OCR and indexing Read-only Ask results

Word Document Ingest

See how a Word document becomes searchable in MemoRay.

This example shows a .docx file queued for local processing, processed on the device, and prepared for search and grounded Ask results.

01 Staged

MemoRay early access screenshot showing a staged Word document queued locally for ingest.

Queued locally before processing.

Review selected files before processing starts.

02 Processing MemoRay early access screenshot showing a Word document being processed locally during ingest.

Word text is processed on this device.

Word structure and text are processed locally.

03 Indexed MemoRay early access screenshot showing a Word document fully indexed and searchable in the app.

Ready to search and ask.

After indexing, the document can appear in search and grounded Ask results.

How MemoRay Is Different

Built for private document memory — not generic AI chat.

MemoRay is not trying to replace every AI assistant, file search tool, or note app. It focuses on one job: helping you search and ask across selected personal records while keeping the core workflow local and source-visible.

Alternative

What it does well

Where MemoRay is different

Cloud AI assistants

Powerful general reasoning and fluent answers.

MemoRay is designed for private document memory where core workflows do not require uploading your files.

Local AI chat tools

Run models locally and experiment with private inference.

MemoRay focuses on document ingestion, retrieval, evidence, and user-controlled memory — not just chatting with a model.

File search / desktop search

Find filenames and text matches.

MemoRay adds natural-language retrieval and evidence-backed Ask behavior across imported records.

Cloud note/document apps

Collaboration, sync, and convenience.

MemoRay starts from local control, no required login, and no required document upload.

Privacy & Local Processing

Trust is designed into the workflow, not added as a slogan.

MemoRay's early-access product is intentionally explicit about what runs locally, what is optional, and what is still being hardened. The goal is not vague privacy language - it is clear, inspectable product behavior.

No Upload

Core MemoRay document processing does not require file upload.

No document content is uploaded for the core local workflow.

Optional Telemetry

Anonymous usage sharing is off by default.

Users can choose to share anonymous usage summaries. Documents, queries, embeddings, and file names are excluded.

Local Processing

Core document workflows are designed to run on the user's device.

Ingestion, OCR or text extraction, indexing, search, and evidence retrieval operate locally for the core path.

Storage Note

MemoRay is careful about local storage claims.

Documents and blobs use a password-derived encrypted local storage path, while metadata, text, and chunks still rely on SQLite. Full encryption parity remains a product-hardening item.

Trust Architecture

The core MemoRay workflow is designed around local processing.

Files move through local processing, local retrieval, and evidence-backed Ask behavior without requiring document upload for the core workflow.

Your files remain the source of truth for search and Ask results.
OCR, extraction, indexing, lexical search, and local vector retrieval run locally.
Read the privacy policy and FAQ for more detail on local processing and current product limits.

Early Access

Early access is opening in focused stages.

Join the waitlist for the first Windows early-access cohort while onboarding remains limited and hands-on.

Early Access Limited

First early-access build with the core local workflow.

The first Windows build is focused on import, OCR, indexing, search, and grounded Ask results.

PDF, DOCX, and image import
Local OCR and search
Evidence-backed Ask results

Next In Progress

Founder Access onboarding and delivery.

The next step is hands-on cohort access, installer delivery, and recovery flows.

Founder Access enrollment for the early cohort
Installer delivery and download recovery
License and download recovery messaging

Later Later

Broader platform expansion.

MemoRay starts with Windows early access. Future platform priorities will be shaped by early user demand.

macOS and additional desktop platforms
Mobile apps for iOS and Android
Opt-in sync and team features