Everything You Need, Nothing You Don't
SeekerPDF combines AI-powered analysis, professional editing, smart scanning, and enterprise security into one beautifully designed iOS application.
7-Layer AI Guardrails
Enterprise-grade trustworthy AI
Every LLM API call passes through a defense-in-depth middleware pipeline. Input validation, prompt injection detection, PII auto-redaction, content policy enforcement, output validation, PII leak prevention, and hallucination risk scoring — with full observability and audit logging.
Key Capabilities
- Input validation: length limits, format checks, empty input rejection
- Rate limiting: 30 requests/min, 500/hour (configurable)
- Prompt injection detection: 17 weighted regex patterns with confidence scoring
- PII redaction: 12 types including credit cards with Luhn validation
- Content policy: blocks weapons, hacking, self-harm, CSAM content
- Output PII leak prevention: scans AI responses before delivery
- System prompt leak detection via n-gram window matching
- Hallucination risk scoring in AI responses
- Full observability: injection scores, toxicity trends, audit trail
- 3 config presets: Default, Strict, Permissive
Advanced RAG Pipeline
5 chunking strategies + 7 retrieval methods
Industry-leading Retrieval-Augmented Generation with intelligent document chunking and multi-strategy retrieval. GraphRAG discovers entity relationships, Corrective RAG self-corrects queries, and Hybrid Search fuses keyword + semantic results.
Key Capabilities
- Semantic chunking with NLTokenizer sentence overlap
- Structural chunking via heading detection (Markdown, CHAPTER, numbered)
- Agentic chunking: LLM-driven paragraph grouping by coherence
- Hierarchical chunking: 4-level doc → section → paragraph → sentence tree
- Hybrid Search: BM25 + TF-IDF with Reciprocal Rank Fusion
- Re-Ranking: LLM scores each (query, chunk) on 0-10 relevance
- Corrective RAG: auto-reformulates queries when retrieval quality is low
- GraphRAG: entity extraction → co-occurrence graph → traversal
- Contextual Retrieval: LLM-generated context per chunk before indexing
- ChunkingRouter auto-selects optimal strategy per document
Chat with Documents (RAG Q&A)
5 query modes with source citations
Have natural conversations with any PDF using RAG-powered answers with source citations. Voice input, auto-read responses, conversation memory, and 5 specialized query modes.
Key Capabilities
- Natural language question answering with source citations
- 5 modes: Standard, Deep Analysis, Fact Check, Creative, Summary
- Conversation memory for contextual follow-ups
- Voice input (speech-to-text) for hands-free queries
- Auto-read responses aloud with premium TTS
- Multi-document cross-referencing
- Suggested prompts for quick insights
- Export and share conversations
Document Translation
20+ languages with quality scoring
Translate entire documents across 20+ languages using a 4-agent pipeline: language detection, translation, quality scoring, and automatic refinement. Preserves meaning, tone, and formatting.
Key Capabilities
- 20+ languages including RTL (Arabic, Hebrew) support
- Language Detection Agent identifies source language and formality
- Translation Agent preserves meaning, tone, sentiment, and context
- Quality Agent scores: meaning, tone, grammar, coherence, cultural fit
- Refinement Agent auto-improves if quality < 0.75 threshold
- Scan & Translate: camera → OCR → translate pipeline
- Paragraph-boundary chunking for large documents
- Translation history with export support
Multi-Agent Orchestration
Parallel execution with error isolation
The AgentOrchestrator coordinates specialized AI agents with goal-based task selection, parallel execution, dependency resolution, and aggregated results. Individual agent failures are isolated — the pipeline continues.
Key Capabilities
- 6+ specialized agents: Summary, Analysis, QA, Review, Translation, Extraction
- Goal-based agent selection: Quick Summary, Full Analysis, etc.
- Parallel task execution with dependency resolution
- Error isolation: agent failures don't crash the pipeline
- Aggregated results with confidence scoring
- Agent output caching (LRU with TTL) for repeated queries
- Evaluator-optimizer loop for quality improvement
Smart OCR Scanner
Camera to searchable PDF with AI
Transform physical documents into intelligent digital assets. Vision-powered OCR with auto-edge detection, batch scanning, image filters, and instant scan-to-translate pipeline.
Key Capabilities
- Camera-based scanning with auto-edge detection
- Apple Vision framework OCR with high accuracy
- Batch scanning for multi-page documents
- Auto deskew and perspective correction
- Brightness, contrast, and filter adjustments
- Scan → OCR → Translate pipeline
- Save as searchable PDF with instant AI analysis
Professional PDF Editor
Full annotation and signature toolkit
Complete PDF editing with highlights, text annotations, freehand signatures (PencilKit), form filling, stamps, redaction, and page management — all native on iOS.
Key Capabilities
- Highlight, underline, and strikethrough text
- Freehand signatures with PencilKit (Apple Pencil + finger)
- Signature library: save and reuse multiple signatures
- Text annotations and sticky notes
- Professional stamps: APPROVED, CONFIDENTIAL, DRAFT
- Interactive PDF form field filling
- Redaction for permanent information removal
- Page crop, rotate, and reorder
Book Summarization
Chapter-by-chapter with page-level analysis
Multi-mode book and note summarization: extract chapters, analyze page-by-page, run AI agents, and synthesize insights — with a live progress pipeline UI showing each step.
Key Capabilities
- Note mode (fast, 2-3 LLM calls) for small/medium docs
- Book mode (deep, multi-stage) for large documents
- Chapter detection and individual chapter summaries
- Page-level analysis for detailed content extraction
- AI agent integration for themes, entities, and key insights
- Live progress UI with step-by-step status indicators
- Finalized summary with export and sharing
Article Peer Review
Academic-style review with scoring
Specialized analysis for research papers and journal articles. Methodology evaluation, critical review, findings extraction, and learning enhancements — mimicking academic peer review.
Key Capabilities
- Methodology analysis and evaluation scoring
- Critical review of findings, conclusions, and claims
- Research quality assessment with detailed scoring
- Learning enhancements: simplify, quiz, real-world examples
- 5 result tabs: Summary, Findings, Methods, Critical Analysis, Enhance
- Export and share reviews
PII Redaction & Masking
12 sensitive data types auto-detected
Comprehensive PII detection using regex patterns with confidence scoring. Auto-redact before LLM calls and auto-scan AI responses for leaks. Partial masking (e.g., jo***@example.com) available.
Key Capabilities
- Email addresses (regex + validation)
- Phone numbers (multiple formats: US, international)
- Social Security Numbers (SSN)
- Credit cards with Luhn algorithm validation
- IP addresses (IPv4)
- Dates of birth, passport numbers, driver licenses
- Bank account numbers and medical record IDs
- Partial masking mode for readable redaction
- Input redaction + output leak prevention
Voice I/O & Audio Reader
Speech-to-text + premium text-to-speech
Full voice integration: speech-to-text for all query fields, premium human-like TTS with adjustable rate and pitch, and a complete audio reader with playback controls.
Key Capabilities
- Speech-to-text input via SFSpeechRecognizer
- Premium text-to-speech with AVSpeechSynthesizer
- Adjustable playback speed (0.5x to 2x) and pitch
- Multiple voice options and languages
- Background playback (continues when minimized)
- Full audio reader with play/pause/skip controls
File Manager & Cloud Sync
Organize with iCloud, Google Drive, Dropbox
Complete document management with local and cloud storage. Browse, search, sort, convert formats, and share via any app. Remote storage integration for iCloud, Google Drive, and Dropbox.
Key Capabilities
- Browse and organize documents locally
- iCloud, Google Drive, Dropbox cloud sync
- Search files by name and content
- Sort by name, date, or size
- Convert between PDF, Word, TXT, PPTX
- Share via AirDrop, email, and other apps
- Recent files quick access (up to 20)
- Bookmarks (up to 100 per document)
Ready to Get Started?
Download SeekerPDF and unlock the full power of AI-driven document intelligence.