Lam is a one-of-a-kind, question-focused, browser-based diagnostic system that thinks, reasons, and evolves like a top-tier systems engineer. Built entirely using HTML, CSS, and JavaScript — no backend, no external APIs — Lam simulates human-like troubleshooting intelligence to deliver highly probable, explainable PC problem diagnoses.
A core module that simulates metacognitive problem-solving:
- Adaptive questioning using Bayesian inference and fuzzy logic
- Confidence scoring that updates in real time
- Multisymptom correlation (e.g., “freezes + noises + Event ID 51”)
Switch between expert personas for domain-specific insight:
- Hardware Hermit – Power, thermal, and physical-level faults
- OS Oracle – Deep system internals, drivers, event logs
- Network Ninja – IP stack failures, DNS, adapter issues
- Forensics Falcon – Malware, behavioral anomalies, stealth issues
Each persona customizes the diagnostic path, questioning tone, and logic hierarchy.
A unique interface inspired by crime-solving timelines:
- Live-updating diagnosis probability graphs
- Breadcrumb trail of user responses and Lam’s interpretations
- “What If” simulation mode for alternate scenario modeling
- Visual explanation of every decision and next step
- Clarifying sub-questions based on ambiguity detection
- Emotion-aware UI cues (frustration, urgency, confusion)
- Natural language pattern recognition (e.g., “it acts weird when I open Chrome”)
- Deep transparency: ask Lam “Why this question?” at any point
Lam runs 100% client-side:
- Fully functional offline as a Progressive Web App
- Uses Service Workers to cache diagnostic models and UI assets
- LocalStorage saves session history, user decisions, and confidence shifts
- Sessions exportable as JSON or PDF for technicians
Peek inside Lam’s “mind”:
- Real-time logic tree with weighted decisions
- Confidence heatmap of possible issues
- Internal monologue simulator (“Based on the PSU age and error timeline…”)
- Self-corrective reasoning: changes path if assumptions fail
- HTML5 — Clean semantic structure
- TailwindCSS — Apple-level minimalist UI/UX
- Vanilla JavaScript (ES6+) — Modular expert systems, decision trees, and inference logic
- JSON — Diagnostic knowledge base, persona modules, and logic graphs
- Service Workers — Full offline PWA support
- LocalStorage — State persistence and historical analysis
- 🔌 Plugin system to add community-made diagnosis modules
- 🌐 RTL interface support for Arabic and other languages
- 🔍 Smart scanner (experimental): real-time system feedback parser using Web APIs
- 🧠 Learn Mode: teaches users basic troubleshooting as they answer questions
Lam is not just a tool — it’s a digital diagnostician.
It merges the logical rigor of a technician, the curiosity of a doctor, and the design elegance of a creative visionary — all running in your browser.
No installation needed. Just open index.html in your browser.
To install Lam as a PWA:
Click the install icon in your browser (available in Chrome, Edge, etc.) for offline access.
Inspired by the genius of cognitive systems, cyber forensics, and human intuition.
MIT License – Feel free to use, remix, or enhance Lam for your own projects.