Home

Welcome to the Official WishPilot Wiki

Welcome to the comprehensive technical documentation and user manual for WishPilot — the high-performance, local-first stealth interview copilot and real-time speech intelligence engine.

Designed, architected, and maintained by Vishwjeet Singh Vilkhu.

---

What is WishPilot?

WishPilot is an advanced desktop application engineered for candidates preparing for and navigating high-stakes technical, operational, and leadership interviews. Built with Electron, React 19, and Vite, it bridges real-time voice intelligence, multi-stream prompt adaptation, and zero-distraction desktop overlay technology into a single, cohesive workflow.

Core Philosophy

1. Local-First & Zero Telemetry: No audio, transcripts, resume text, or API credentials ever touch a WishPilot central server. All data resides in encrypted local storage.
2. Bring Your Own Key (BYOK): Connect directly to official provider APIs (Groq, Cerebras, Together AI, Fireworks AI, NVIDIA NIM, Hugging Face, OpenRouter, OpenAI, and Google Gemini) over direct HTTPS.
3. Sub-Second Latency: Audio DSP pipelines buffer and transcribe speech in 180ms to 350ms using Groq Whisper Large v3 Turbo, allowing candidates to scan answers with zero awkward silence.
4. Adaptive Multi-Industry Intelligence: Beyond software engineering, WishPilot natively adapts to 9 distinct corporate and clinical domains with tailored frameworks.

---

Documentation Index

Explore the detailed documentation chapters below:

Chapter	Title	Focus Area
01	Getting Started & Installation	System prerequisites, running standalone binaries, and building from source.
02	Architecture & Core Engines	Electron multi-window lifecycle, Web Audio DSP pipeline, and display affinity.
03	Supported AI Providers & Setup	Setup instructions, API keys, models, and latency benchmarks for all 9 providers.
04	Multi-Industry Category Engine	Frameworks (LAST, 3-Statement, BANT, STAR, CIRCLES) and prompts for 9 streams.
05	Instant Answer Refinement Pills	Real-time answer compression, deep technical injection, case study grounding, and simplification.
06	Stealth HUD & Hotkeys Reference	Overlay modes (Studio, HUD, Notch), screen capture, click-through, and shortcuts.
07	Session Management & Debrief	Organizing interviews by company/role, tracking transcripts, and exporting post-call debriefs.
08	Troubleshooting & FAQ	Common audio routing, rate limiting, and display capture resolution steps.

---

Quick Feature Overview

                          ┌────────────────────────┐
                          │   Live Audio / Speech  │
                          └───────────┬────────────┘
                                      │ (Web Audio DSP / 16kHz WAV)
                                      ▼
                          ┌────────────────────────┐
                          │  Groq Whisper Turbo    │  <-- ~250ms STT
                          └───────────┬────────────┘
                                      │ Real-Time Transcript
                                      ▼
       ┌─────────────────────────────────────────────────────────────┐
       │              WishPilot Core Orchestration                   │
       │  - Active Category Domain (IT, BPO, Finance, PM, HR, etc.)  │
       │  - Candidate Background & Resume Alignment                  │
       │  - Vision Screen Context (LeetCode / Diagrams)              │
       └──────────────────────────────┬──────────────────────────────┘
                                      │ Direct HTTPS (BYOK)
                                      ▼
       ┌─────────────────────────────────────────────────────────────┐
       │     Unified Inference Engine (9 Supported Providers)        │
       │  Groq • Cerebras • Together • Fireworks • NVIDIA • OpenAI   │
       └──────────────────────────────┬──────────────────────────────┘
                                      │ Word-by-Word Stream
                                      ▼
       ┌─────────────────────────────────────────────────────────────┐
       │              Candidate Floating HUD / Notch                 │
       │  - 10-Second TL;DR Spoken Punchline                         │
       │  - Structured Spoken Breakdown                              │
       │  - Instant Refinement Pills (Shorter, Technical, etc.)      │
       └─────────────────────────────────────────────────────────────┘

---

Legal & Intellectual Property Notice

WishPilot is an open-source project licensed under the GNU General Public License v3.0 (GPL-3.0-or-later).

• Author: Vishwjeet Singh Vilkhu (@vishwjeet27)
• Repository: https://github.com/vishwjeet27/wishpilot
• Academic / Software Citation: Please refer to CITATION.cff when referencing WishPilot.

All derivative works or redistributions must maintain original author attribution and remain strictly open-source under GPL-3.0.