Release v0.0.2

Intervity v0.0.2

Complete Subject Mastery (Zero-to-Hero QBL), 32+ Node.js Deep Architectural Questions, Interactive 3-Tier Progressive Hints, and Precision Studio UI Edition.

Highlights & New Features in v0.0.2

• Zero-to-Hero Complete Subject Mastery (QBLEngine & QBLCurriculumGenerator): Support for both 5-chapter flagship roadmaps and full 20-subtopic Zero-to-Hero multi-chapter curricula across all 6 cognitive levels of Bloom's Taxonomy (Remember -> Create).
• Expanded Node.js Domain Knowledge Base (NodeJSDomain): 32 authoritative questions spanning Event Loop & libuv phases, Reactor pattern (epoll/kqueue), V8 generational GC & hidden classes, 8KB FastBuffer slab allocator, Stream backpressure (highWaterMark, stream.pipeline, for await...of), worker_threads + SharedArrayBuffer, SO_REUSEPORT vs IPC socket handover, AsyncLocalStorage, ReDoS, and Prototype Pollution defense.
• Zero Cross-Concept & Cross-Subtopic Duplication: Subtopic-aware question prioritization, multi-concept procedural question expansion, and SQLite turn history hydration on session resume.
• Interactive 3-Tier Progressive Hints & Source Citations: In-card progressive hint reveal (Conceptual Nudge -> Mechanism Clue -> Elimination Guide), Learning Objective banners, Bloom cognitive badges, and authoritative specification references.
• Agent Coaching Harness QBL Synchronization: Automatic synchronization of completed QBL subjects and subtopics into persistent UserCareerMemory and Mascot XP progression, plus new Node.js runtime analogies in AnalogicalTeachingStrategy.
• Studio Dark UI Hierarchy: Precision obsidian/zinc dark surface theme across all 10 screens and core components.

Included Binaries

• Android APK: Intervity-v0.0.2.apk (242 MB)
• iOS IPA: Intervity-v0.0.2.ipa (14 MB, Bundle ID: help.dina.intervity, Team: BXS665PPF3)

Release v0.0.1

Intervity v0.0.1

Initial release of Intervity — an on-device, offline AI technical interview coach and question-driven learning (QBL) application.

Highlights & Features

• On-Device LLM & Speech: Fully offline inference utilizing GGUF models via llama.rn and speech-to-text with Whisper/Sherpa ONNX.
• Pedagogical QBL Engine: Multi-tiered technical question catalog across software engineering domains, Bloom taxonomy, and real-time Star Method evaluation.
• Teddy Mascot Companion: Adaptive coaching personality with XP progression and empathetic relationship cues.
• Full-Duplex Voice & Instant Barge-In: Real-time microphone listening stream with sub-15ms voice interruption and playback flushing.
• Voice Profile Biometrics: 192-dimensional acoustic embedding extraction for speaker verification and personalized candidate interaction.
• Offline Persistence: Local SQLite storage via @op-engineering/op-sqlite.

Included Binaries

• Android APK: Intervity-v0.0.1.apk (242 MB)
• iOS IPA: Intervity-v0.0.1.ipa (14 MB, Bundle ID: help.dina.intervity, Team: BXS665PPF3)