luje.io Trainer helps you move from pilot to production. Timed AI PM scenarios, persona-aware probing, and real LLM scoring - Deploy. Monitor. Train.
session preview - live statsLIVE
Session stats- avg • 0 questions
Avg score-last session
Weak area-auto-surfaces
Bank70scenarios
Progress
Persistence note - Bank is now 70 scenarios + 6 cases + your logged questions - answers live in this browser's localStorage only. No cloud sync. Export regularly (sidebar → Export).
Train - Real AI PM Scenarios, Coach Scoring
One question at a time. Timed. Persona-aware. The trainer remembers your last answers and probes where you're vague - like a real panel.
Session -
session - AI PM Coach⏱ 00:00
---
time to answer
Press “Start” to get your first scenario.
The trainer will tailor the first question to your weak areas (or pull a fresh open-source scenario if you’re new).
browser fallbackSTT: Groq Whisper (good) via hidden luje.io key - no paste needed, or add your own gsk_... TTS: Coqui/ElevenLabs for natural voice, otherwise browser (robotic).
Tip: Aim for 90–150 words, name a framework, quantify success.
Coaching Feedback-
Session Memory
Trainer holds context across turns - adapts tone, references past weak spots.
Every answer is scored on structure, prioritization, stakeholder awareness, technical depth, and communication - with a short drill and a follow-up question, just like a real panel.
Structured feedbackPersonalized drills
USE CASES
AI workflows built for how operators work.
Buyers don't buy "AI strategy" - they buy a specific outcome. Here's what production-grade looks like across the industries and workflows we know best. Tap any card to train on it.
DeployMonitorTrain8 use cases · linked to 70 trainer scenarios
Have a specific workflow in mind?
Tell us the workflow, tool, and implementation approach for your operation.
Case Study Mode - Multi-dimension Scored
Long-form: design, metrics, tradeoffs, stakeholder objection. Scored across all competencies, not just one Q&A.
Review & Transcripts - Your answer next to coaching
Progress Over Time
-Overall avg
-Best competency
-Focus area
Session Reports
Competency Breakdown
Resources Hub - Filtered to Your Weak Areas
Curated open-source frameworks, docs, and readings. Auto-sorted to your two weakest competencies first.
Resume & Portfolio Drafter
Uses your strengths to draft bullets. Edit and export as Markdown/PDF.
Case-Study Portfolio
One-page case template from your Case Study Mode submissions.
Job Application Tracker - Feeds Your Training Bank
Log roles + real interview questions you encountered. Logged questions are auto-tagged and added to your trainer bank (open-source loop).
SETTINGS
Your workspace
Data stays in this browser. Back up with Export, restore with Import. Everything else is optional.
Your data
Private and local. No cloud sync. Export after big sessions.
Downloads luje-trainer-export.json
Tip: Training works with no key (heuristic). Add a Groq key below for real LLM scoring.
DEVDeveloper optionsLLM & voice wiring - hidden from trainees
AI System
Choose the scoring brain. Trainees never need to touch this.
Mock coach active - add Ollama or key to upgrade. No data leaves your browser in mock mode.
Voice Pipeline - STT & TTS
Independent from scoring. Groq key = good mic (STT), not speaker. Good speaker = Coqui (local) or ElevenLabs (cloud).