Build the future of
offline intelligence.

Senior ML Engineer — Scoring & Ranking

📍 San FranciscoFull-time

ML

→

Own VaultBook's multi-field weighted scoring engine — the system that ranks QA search results across 7 content layers (titles at weight 8, labels at 6, OCR text at 5, body at 4, sections at 3, main attachments at 2, section attachments at 1). You'll optimize the scoring pipeline, integrate persistent vote-based reranking (+1M/−1M from userVotes), and ship improvements that users feel in every search.

What you'll do

• Tune and extend the weighted scoring model across all 7 content layers
• Build the persistent reranking layer driven by upvote/downvote signals
• Optimize QA search latency for libraries with 10,000+ entries
• Design A/B frameworks to measure relevance improvements

What you bring

• 4+ years building search ranking or recommendation systems
• Strong background in information retrieval and scoring models
• Experience with client-side ML (TensorFlow.js, ONNX Runtime Web, or similar)
• Comfort with JavaScript performance profiling at scale

JavaScriptTF.jsBM25Learning-to-Rank

Apply for this role →

NLP Research Engineer

📍 San FranciscoFull-time

NLP

→

Design and improve the natural-language query pipeline that lets users ask their library questions in plain English. You'll work on query understanding, intent classification, and result extraction — all running client-side in the browser with zero server calls.

What you'll do

• Build and refine the QA mode query parser and intent detection
• Improve answer extraction and snippet highlighting from search results
• Research lightweight NLP models that fit browser runtime constraints
• Develop query expansion and synonym handling for better recall

What you bring

• MS/PhD in NLP, computational linguistics, or related field
• Experience deploying language models in resource-constrained environments
• Familiarity with tokenization, embeddings, and extractive QA architectures
• Published research or strong open-source contributions in NLP

NLPONNXTransformersWebAssembly

Apply for this role →

Recommendation Systems Engineer

📍 San FranciscoFull-time

ML

→

Build the Related Entries engine — VaultBook's contextual similarity system that surfaces notes users didn't know were connected. The system uses content similarity scoring with a Reddit-style upvote/downvote feedback loop that persistently reshapes recommendations over time.

What you'll do

• Design and improve contextual similarity algorithms for Related Entries
• Integrate vote-based feedback into the recommendation model
• Build cold-start strategies for new entries with limited interaction data
• Optimize recommendation latency for real-time display on entry open

What you bring

• 3+ years in recommendation or personalization systems
• Experience with collaborative filtering, content-based, and hybrid approaches
• Comfort building ML pipelines that run entirely in JavaScript
• Strong intuition for balancing exploration vs. exploitation

Cosine SimilarityTF-IDFJavaScriptIndexedDB

Apply for this role →

Applied AI Researcher — Vote-Based Learning

📍 San FranciscoFull-time

ML

→

Research and develop the persistent vote-based learning system at the heart of VaultBook's personalization. Every upvote (+1M) and downvote (−1M) is stored in the userVotes schema and reshapes both search results and related entry ordering across sessions — a reinforcement signal built from human feedback.

What you'll do

• Design reward models from sparse binary vote signals
• Research how vote decay, recency weighting, and context affect ranking quality
• Build offline evaluation frameworks to measure learning effectiveness
• Publish findings and contribute to the local-first ML research community

What you bring

• MS/PhD with research in reinforcement learning, RLHF, or bandit algorithms
• Experience with preference learning from implicit or explicit feedback
• Strong mathematical foundations in optimization and probability
• Ability to translate research into production JavaScript

RLHFBanditsJSON SchemaJavaScript

Apply for this role →

Computer Vision Engineer — OCR Pipeline

📍 San FranciscoFull-time

CVOCR

→

Own VaultBook's inline OCR engine — automatic text extraction from images embedded in entries, cached per-item, with background warm-up for the top 12 QA results. You'll also extend cross-format OCR that reaches inside DOCX (word/media), XLSX (xl/media), and ZIP archives to extract text from images embedded deep inside compound documents.

What you'll do

• Improve OCR accuracy and speed for handwritten and low-contrast images
• Build the background warm-up pipeline that pre-OCRs top search results
• Extend OCR into compound formats (DOCX media, XLSX media, nested ZIPs)
• Implement per-item OCR caching with invalidation on attachment changes

What you bring

• 3+ years in OCR, document understanding, or computer vision
• Experience with Tesseract.js, ONNX vision models, or browser-based CV
• Understanding of image preprocessing (binarization, deskew, noise removal)
• Comfort with Web Workers and off-main-thread processing

Tesseract.jsCanvas APIWeb WorkersONNX

Apply for this role →

ML Infrastructure Engineer — Client-Side Models

📍 San FranciscoFull-time

MLWebGPU

→

Ensure all AI scoring, similarity calculations, OCR, and suggestions run smoothly in-browser — zero server calls. You'll optimize model loading, manage memory for large libraries, and prepare the infrastructure for WebGPU and WebAssembly acceleration of larger local models.

What you'll do

• Profile and optimize client-side ML model loading and inference
• Build model quantization and compression pipelines for browser delivery
• Implement WebGPU acceleration paths for compute-heavy operations
• Design memory management strategies for devices with limited RAM

What you bring

• Deep experience with TensorFlow.js, ONNX Runtime Web, or WebNN
• Understanding of model quantization (int8, float16) for edge deployment
• Familiarity with WebGPU compute shaders and WGSL
• JavaScript performance optimization at the memory-layout level

TF.jsWebGPUWASMONNX Runtime

Apply for this role →

Embedding Systems Engineer

📍 San FranciscoFull-time

ML

→

Generate, store, and index the content embeddings that drive similarity scoring for Related Entries and the AI Suggestions pager. You'll work on efficient vector operations in the browser, incremental index updates, and embedding models optimized for the VaultBook content domain.

What you'll do

• Select, fine-tune, and deploy embedding models for browser inference
• Build incremental vector index updates as entries are created/edited
• Optimize nearest-neighbor search for real-time Related Entries
• Benchmark embedding quality across diverse note content types

What you bring

• Experience with sentence embeddings (SBERT, E5, or similar)
• Familiarity with approximate nearest neighbor algorithms (HNSW, IVF)
• Ability to compress and quantize models for in-browser use
• Strong JavaScript and TypeScript skills

EmbeddingsANNONNXFloat16

Apply for this role →

Senior Search Engineer — Multi-Field Scoring

📍 San FranciscoFull-time

Search

→

Build and extend VaultBook's search infrastructure — typeahead with real-time dropdown, query history suggestions, paginated results, and the warm-up system that pre-loads text for the top 12 results. Every operation runs client-side with sub-100ms response targets.

What you'll do

• Optimize the multi-field search index for 10,000+ entry libraries
• Implement typeahead with prefix matching and frequency-weighted suggestions
• Build the warm-up pipeline that pre-loads OCR text for top results
• Design search analytics to surface scoring quality issues

Inverted IndexJavaScriptWeb WorkersBM25

Apply for this role →

Deep Indexing Engineer — Document Parsing

📍 San FranciscoFull-time

IndexingJSZip

→

Own the deep file indexing pipeline that extracts searchable text from XLSX (via SheetJS), PPTX (via JSZip XML parsing), PDF (via pdf.js), ZIP archives, and MSG email files. You'll ensure every attached document becomes part of the user's searchable knowledge graph.

What you'll do

• Extend document parsers for new file formats and edge cases
• Build the background extraction pipeline with progress tracking
• Handle nested archives (ZIP within ZIP) and compound documents
• Optimize memory usage for large files processed in the browser

JSZipSheetJSpdf.jsBlob API

Apply for this role →

Semantic Search Specialist

📍 San FranciscoFull-time

NLPSearch

→

Move VaultBook's search beyond keyword matching toward true semantic understanding. You'll develop smart label suggestions through content analysis, build synonym expansion, and lay the groundwork for meaning-based retrieval — all running locally without external API calls.

What you'll do

• Build semantic similarity search alongside the existing keyword engine
• Develop the smart label suggestion system using content analysis
• Implement query intent classification for better result routing
• Research and prototype meaning-based connections between entries

Sentence EmbeddingsONNXTF-IDFJavaScript

Apply for this role →

Ranking & Relevance Engineer

📍 San FranciscoFull-time

SearchML

→

Fine-tune the multi-field scoring weights and vote-based reranking that float the best QA results to the top. You'll work at the intersection of search engineering and machine learning, turning raw relevance scores into results that feel right to the user.

What you'll do

• Run experiments on field weight configurations across content types
• Build evaluation datasets for offline relevance measurement
• Integrate vote signals and recency into the final ranking formula
• Instrument search to collect anonymized quality metrics locally

Learning-to-RankBM25nDCGJavaScript

Apply for this role →

Cryptography Engineer — AES-256-GCM / PBKDF2

📍 San FranciscoFull-time

Crypto

→

Own VaultBook's per-entry encryption — AES-256-GCM with PBKDF2 key derivation at 100,000 iterations, random salt and IV per encrypt cycle. You'll also build the session password cache, lock screen blur overlay, and ensure decrypted content is held in memory only and never written to persistent storage.

What you'll do

• Audit and harden the per-entry encryption implementation
• Research migration paths to newer KDFs (Argon2id via WASM)
• Build key rotation workflows without requiring full re-encryption
• Ensure decrypted plaintext never leaks to disk, clipboard history, or swap

Web Crypto APIAES-GCMPBKDF2Argon2

Apply for this role →

Privacy Engineer — Zero-Telemetry Audit

📍 San FranciscoFull-time

Privacy

→

Audit every code path to guarantee zero network calls — no analytics beacons, no font CDNs, no external requests of any kind. You'll build automated verification tools that flag any outbound request attempt and ensure VaultBook's privacy-by-architecture promise is provably true.

What you'll do

• Build CI checks that fail on any outbound network request in the codebase
• Audit third-party libraries for hidden telemetry or CDN dependencies
• Design a privacy certification process for each release
• Document the zero-telemetry architecture for the security page

CSPDevTools ProtocolStatic AnalysisCI/CD

Apply for this role →

Application Security Engineer

📍 San FranciscoFull-time

Security

→

Secure VaultBook's attack surface — XSS prevention in the rich text editor, safe HTML sanitization for imported content, secure blob URL handling, and protection against local file system injection via the File System Access API.

What you'll do

• Harden the contentEditable editor against XSS and injection attacks
• Audit file import paths (Obsidian import, URL save, attachment upload)
• Build sandboxing for user-generated HTML content in entries
• Perform regular penetration testing and threat modeling

DOMPurifyCSPOWASPFile System Access API

Apply for this role →

Senior Frontend Engineer — Rich Text Editor

📍 San FranciscoFull-time

JScontentEditable

→

Own VaultBook's rich text editor — tables, code blocks, callout blocks, case transforms, headings H1–H6, inline images, font selectors, and text/background color pickers. The editor must handle large entries, paste-from-Word, and maintain clean HTML output across all browsers.

What you'll do

• Extend the editor with new block types and formatting options
• Handle cross-browser contentEditable quirks (cursor, selection, paste)
• Build undo/redo with structural awareness for nested blocks
• Optimize rendering for entries with 100+ images and deep nesting

contentEditableSelection APIMutationObserverVanilla JS

Apply for this role →

Frontend Engineer — File System Access API

📍 San FranciscoFull-time

Web APIs

→

Own VaultBook's storage layer — repository.json, sidecar .md files, the /attachments directory with index.txt manifest, and the autosave system with dirty-flag debouncing and concurrent write guards. You'll push the boundaries of what browser-based file management can do.

What you'll do

• Extend the File System Access API integration for new storage patterns
• Build the deferred cleanup system for orphaned attachments
• Implement conflict resolution for concurrent writes
• Design fallback storage for browsers without FSAA support

File System Access APIIndexedDBBlob APIJSON

Apply for this role →

Frontend Engineer — Canvas Rendering & Charts

📍 San FranciscoFull-time

Canvas

→

Build VaultBook's analytics dashboards — canvas-rendered label pie charts, page utilization graphs, 14-day activity timelines, attachment type breakdowns, and strength metric pills. All visualizations must be smooth, responsive, and render without external charting libraries.

What you'll do

• Design and implement custom Canvas 2D chart components
• Build animated transitions for data updates and filtering
• Ensure charts are accessible with ARIA descriptions and fallback text
• Optimize rendering for libraries with thousands of data points

Canvas 2DrequestAnimationFrameSVGVanilla JS

Apply for this role →

Full-Stack Engineer — Built-in Tools

📍 San FranciscoFull-time

JSTools

→

Ship VaultBook's 14+ built-in tools — File Analyzer, PDF Merge/Split/Compress, MP3 Cutter & Joiner, Password Generator, Folder Analyzer, Threads, RSS Reader, Save URL → Entry, and more. Each tool runs entirely in-browser and integrates seamlessly into the workspace UI.

What you'll do

• Build new tools from concept to production in weekly cycles
• Integrate tools with VaultBook's entry and attachment systems
• Handle complex file processing (audio slicing, PDF operations) client-side
• Design consistent UX patterns across all tools

Web Audio APIpdf-libBlob APIVanilla JS

Apply for this role →

Performance Engineer — Large Library Optimization

📍 San FranciscoFull-time

Perf

→

Make VaultBook fast at scale. You'll optimize pagination, lazy attachment loading, background OCR warm-up, sidebar rendering for 10,000+ entries, and the single-file build that keeps the entire app under aggressive size targets.

What you'll do

• Profile and fix rendering bottlenecks in the entry list and sidebar
• Implement virtual scrolling for large libraries
• Optimize the JSON parse/serialize cycle for repository.json
• Reduce time-to-interactive on cold load with 5,000+ entries

Chrome DevToolsPerformance APIWeb WorkersIntersectionObserver

Apply for this role →

Accessibility Engineer

📍 San FranciscoFull-time

A11y

→

Make VaultBook fully keyboard-navigable and screen-reader friendly. You'll implement ARIA labels and roles across every UI surface, build focus management for modals and sidebars, and ensure the rich text editor is operable without a mouse.

What you'll do

• Audit and fix WCAG 2.1 AA compliance across all views
• Build keyboard navigation for sidebar, modals, and tool panels
• Implement screen reader announcements for dynamic content updates
• Design high-contrast and reduced-motion mode support

ARIAWCAG 2.1axe-coreNVDA/VoiceOver

Apply for this role →

Data Platform Architect — Repository Schema

📍 San FranciscoFull-time

Data

→

Own VaultBook's data model — the repository.json schema that stores items, pages, sections, labels, pagePaths, userVotes, and version history snapshots. You'll design migration strategies, schema evolution, and ensure the JSON structure stays performant as libraries grow.

What you'll do

• Design schema changes that are backward-compatible with existing repos
• Build automated migration scripts for major version upgrades
• Optimize JSON read/write performance for multi-megabyte repository files
• Define data integrity constraints and validation rules

JSON SchemaFile System Access APIIndexedDBMigration

Apply for this role →

Data Pipeline Engineer — Attachment Extraction

📍 San FranciscoFull-time

Pipeline

→

Build the background attachment text extraction pipeline that indexes content from text-based files and images. You'll manage the index.txt manifest, handle re-indexing on attachment changes, and optimize the extraction queue for minimal UI jank.

What you'll do

• Build a priority queue for background text extraction
• Manage the attachment index manifest with delta updates
• Handle file-type detection and router to format-specific extractors
• Optimize memory for processing large files without blocking the UI

Web WorkersBlob APIStreamingFile API

Apply for this role →

Feature Extraction Engineer — SheetJS / pdf.js

📍 San FranciscoFull-time

Parsingpdf.js

→

Parse structured content from PPTX slides via JSZip (reading XML inside the archive), XLSX sheets via SheetJS, and PDF pages via pdf.js. You'll turn opaque binary files into clean, searchable text that feeds VaultBook's deep indexing pipeline.

What you'll do

• Maintain and extend parsers for XLSX, PPTX, PDF, DOCX, and MSG
• Handle edge cases — password-protected files, corrupted archives, huge sheets
• Extract metadata (author, dates, slide notes) alongside content
• Benchmark extraction throughput and optimize for browser limits

SheetJSJSZippdf.jsXML Parsing

Apply for this role →

Data Quality Engineer

📍 San FranciscoFull-time

QA

→

Ensure data integrity across the repository — validate JSON structure, detect orphaned attachments, flag inconsistencies between index.txt and the /attachments directory, and build self-repair routines that fix issues automatically on load.

What you'll do

• Build validation routines that run on every repository load
• Design self-healing for common corruption patterns
• Create integrity dashboards visible in the analytics panel
• Write regression tests for every known data edge case

JSON ValidationFile SystemJestCI/CD

Apply for this role →

Senior Product Designer — Workspace UI

📍 San FranciscoFull-time

Design

→

Shape VaultBook's visual identity — the frosted-glass panels, sidebar toggle mechanics, floating action button, modal stacking, and responsive layout that breaks at 900px. You'll own the design system and ensure every new feature feels native to the workspace.

What you'll do

• Define and maintain the VaultBook design system (tokens, components, patterns)
• Design new features from wireframe through polished spec
• Collaborate daily with frontend engineers on implementation fidelity
• Conduct design critiques and maintain visual consistency across 14+ tools

FigmaCSSDesign SystemsPrototyping

Apply for this role →

UX Researcher

📍 San FranciscoFull-time

UXR

→

Run the usability studies that shape VaultBook's information architecture — the sidebar time tabs (Recent, Due, Expiring, Tools), the search flow, the entry creation modal, and the Kanban board layout. You'll generate insights that directly drive product decisions.

What you'll do

• Plan and conduct moderated usability tests with real users
• Run card-sorting and tree-testing for navigation improvements
• Analyze task completion rates and identify friction points
• Present actionable findings to engineering and design weekly

MazeUserTestingFigmaDovetail

Apply for this role →

Interaction Designer — Micro-Animations

📍 San FranciscoFull-time

Motion

→

Craft the micro-interactions that make VaultBook feel alive — smooth transitions, hover states on cards, close-confirmation dialogs, accordion section expansion, sidebar slide-in, and the floating action button menu. Every animation must be performant, purposeful, and respect reduced-motion preferences.

What you'll do

• Design and implement CSS/JS animations across all UI surfaces
• Build spring-based timing curves for natural-feeling motion
• Ensure all animations degrade gracefully with prefers-reduced-motion
• Prototype interaction concepts in code before design handoff

CSS TransitionsWeb Animations APIrequestAnimationFrameFigma

Apply for this role →

AI Product Manager — Suggestions & QA

📍 San FranciscoFull-time

ProductAI

→

Shape the AI Suggestions carousel and QA search experience. You'll define what each of the 4 suggestion pages shows (upcoming entries, weekday patterns, recent files, recent tools), decide how vote-based learning affects ranking, and set the product direction for VaultBook's intelligence layer.

What you'll do

• Define the product roadmap for AI features across Plus and Pro tiers
• Write specs for new suggestion algorithms and search improvements
• Analyze user feedback from Telegram to prioritize AI enhancements
• Collaborate with ML engineers on feasibility and experimentation

Product StrategyML FamiliarityUser ResearchAnalytics

Apply for this role →

Technical Product Manager — Tools Platform

📍 San FranciscoFull-time

Product

→

Own the roadmap for VaultBook's 14+ built-in tools and decide what ships next. You'll balance user requests, technical feasibility, and product coherence to grow the tools platform without bloating the workspace.

What you'll do

• Prioritize tool development based on user feedback and usage patterns
• Write detailed specs for new tools (PDF, media, data analysis, etc.)
• Define the Plus/Pro tier boundaries for tool access
• Coordinate cross-functional delivery with engineering and design

Product StrategyTechnical WritingRoadmappingUser Research

Apply for this role →

DevOps Engineer — Single-File Build Pipeline

📍 San FranciscoFull-time

CI/CD

→

Manage the build pipeline that compiles VaultBook — all JavaScript, CSS, HTML templates, embedded libraries (SheetJS, JSZip, pdf.js, Tesseract worker) — into a single deployable HTML file with zero external dependencies. Every byte matters.

What you'll do

• Optimize the bundling pipeline for minimal output size
• Build automated smoke tests that verify the single-file output works
• Manage library vendoring and version pinning for embedded dependencies
• Create release automation for versioned builds with changelog generation

Rollup/esbuildGitHub ActionsTerserBase64 Inlining

Apply for this role →

WebAssembly Engineer — Edge Runtime

📍 San FranciscoFull-time

WASM

→

Bring WebAssembly to VaultBook's heaviest workloads — OCR processing, encryption key derivation (Argon2id), PDF rendering, and future ML inference. You'll compile C/Rust libraries to WASM, optimize memory layout, and integrate them seamlessly with VaultBook's JavaScript runtime.

What you'll do

• Compile and optimize WASM modules for OCR and cryptography
• Design the JS↔WASM bridge with minimal serialization overhead
• Profile and reduce WASM memory usage for constrained devices
• Evaluate WebGPU compute as an acceleration path for ML models

EmscriptenRust/wasm-bindgenWebGPUSIMD

Apply for this role →

QA Engineer — Cross-Browser Testing

📍 San FranciscoFull-time

QA

→

Test VaultBook across Chrome, Edge, Arc, and Brave on Mac, Windows, and Linux — every release, every platform, every time. You'll build automated test suites, catch rendering differences, and ensure the rich text editor behaves identically across all Chromium-based browsers.

What you'll do

• Build end-to-end test suites with Playwright across all target browsers
• Create visual regression tests for UI components
• Test File System Access API behavior across OS and browser combinations
• Perform manual exploratory testing before every release

PlaywrightChromiumPercyGitHub Actions

Apply for this role →

Technical Writer

📍 San FranciscoFull-time

Docs

→

Write the storage tutorial, update banners, FAQ, in-app tooltips, release notes, and every piece of text that guides users through VaultBook. You'll also maintain the public documentation site and the security/privacy pages that explain VaultBook's architecture to technical audiences.

What you'll do

• Write clear, concise documentation for every new feature
• Maintain the security, privacy, and terms pages
• Create in-app onboarding flows and contextual help
• Produce release notes and changelog entries for each version

MarkdownJekyllTechnical WritingAPI Docs

Apply for this role →