Presentation Projects
Hello there,
I'm Richard. Welcome to my page!
You can find here small, practical, problem-solving case studies covering some details about automation, automation tools, AI, guides, webpages, and more. For any question, feel free to contact me.
See you.
Projects
Foresight
Deterministic life/career decision engine. Type any life decision — AI maps every possible outcome with real probabilities from 310K+ verified data points. Three knowledge layers: sacred texts, behavioral science, and institutional research. Features: Crash Test, Reverse Engineering, 3D walk-through, photo/voice input, 31 scenario templates.
Next.js 16 · React Flow · Claude Haiku · Supabase pgvector (66K embeddings) · Three.js · Dagre · Web Audio
Live (private deploy) · Demo (soon) · Technical Details
▶ VIDEO COMING SOON
310K+ data points · 114 JSON files · 7 live APIs · 31 templates · Cross-device PIN sync
• RAG with 66K embeddings instead of pure-LLM — grounded predictions, not hallucinations
• AI cascade (3 providers) for resilience — if one fails, next picks up
• Deterministic seeded randomness — same input always produces same tree
• $0/mo infrastructure — Supabase free tier + Vercel edge
AI Legal Motions Generator
AI-powered document automation: upload a PDF, extract case data, generate completed legal motions in seconds.
Python · Claude AI · pdfplumber · Streamlit
Watch Demo · Technical Details
PDF parsing · Structured data extraction · Court-ready document generation · Zero-friction UI
• pdfplumber over PyPDF2 — handles complex multi-column PDF layouts
• Claude for structured parsing — extracts 23+ data fields from messy legal text
• PDF to completed motion in under 10 seconds, 90%+ format coverage
Machine Learning Image Recognition Desktop App
Machine-learning image recognition app that finds your photos by keyword. After the initial index and with your preferred model (I implemented more than one), you can search images by tags, text (OCR), visual similarity, date, or folder filters.
It is fully offline and private.
Python · CLIP · BLIP · EasyOCR · FAISS · PyQt6 · SQLite
Demo (soon) · Technical Details
▶ VIDEO COMING SOON
Multi-model indexing · Offline-first · Tag/OCR/similarity search · Cross-platform desktop GUI
• CLIP + BLIP for semantic understanding — search by meaning, not just filename
• FAISS for instant vector search across 100K+ indexed images
• Fully offline — no cloud, no API calls, all data stays local
Deal Alert Engine
Automated e-commerce price monitoring system. Tracks product watchlists on Shopee, detects price drops, sends real-time Telegram alerts. Scheduled scraping with browser automation, auto file processing with Python, premium Google Sheets dashboard.
Google Apps Script · Python · UIVision · Telegram Bot API
Currently inactive · Demo (soon) · Technical Details
Multi-marketplace scraping · Price history tracking · Configurable deal filters · Instant Telegram alerts
• Google Apps Script for free, reliable cron scheduling
• UIVision browser automation for JS-rendered marketplace pages
• Scan → filter → alert pipeline in under 90 seconds, $0/mo
Use Cases
- RAG systems – 66K embeddings + pgvector at $0/mo
- Web scraping at scale – Playwright + Apify across multiple marketplaces
- Video automation – Voice/text prompt to video (Remotion)
- ELO prediction models – XGBoost for sports (70%+ accuracy)
- AI legal document automation – PDF to court-ready motions (pdfplumber + Claude)
- Voice input processing – Whisper transcription to simulation scenarios
- Excel reporting automation – openpyxl generation from trade/event logs
- AWS EC2 bot deployment – Python bots with systemd (24/7)
- Smart contract integration – CTF conditional tokens + proxy wallet transactions
- Web creation – VS Code, Claude Code, Codex CLI
Others
- Life/career simulation – deterministic decision trees + sacred texts
- 3D visualization – Three.js force-directed graphs + walk-through
- Crash Test – headless multi-scenario comparison
- Deal alert engine – Google Sheets + Telegram price-drop automation
- Photo-to-simulation – Claude Vision + decision tree generation
- Reverse engineering – trace outcomes back to root with compound probability
- Knowledge graph – 137 keywords, 6 clusters, Three.js 3D
- Multi-source scraping – Indonesia rental + Perth rental + e-commerce
- Google Ads + Meta Ads – campaign setup and optimization
- Cross-device sync – Vercel KV PIN system for mobile/desktop
- LaunchAgent scheduling – macOS automated tasks with launchctl + plist
- Email alerting – smtplib + Gmail SMTP for automated notifications
- ML image search – CLIP + BLIP + FAISS offline image recognition desktop app
- Instant search – Typesense full-text search with typo tolerance
- Newsletter publishing – Substack content pipeline
- Product analytics – PostHog session recording + event tracking
- CI/CD pipelines – GitHub Actions for automated testing and deploy
- Desktop GUI – PyQt6 native app for ML image recognition
- AI voice – ElevenLabs text-to-speech for content generation
- Ultra-fast inference – Groq LPU for sub-second LLM responses
Technical Writing
Execution Notes
Execution notes documenting some step-by-step I went through and system implementation.
→ Read guides
Technical Writing
Step-by-step technical guides from real projects.
View all guides
- How I Built a RAG System with 66K Embeddings at $0/mo
- Building Foresight: A Deterministic Path Simulator
- 3D Force-Directed Graphs with Three.js from Scratch
- Crash Test: Comparing Life Paths with Headless Simulation
Experiments & Research
Side explorations — tools and prototypes I built to learn, test ideas, or solve personal problems.
Cafepedia.id — Cafe discovery platform, 787K+ places indexed (Indonesia)
Knowledge Base World — 3D knowledge graph, 180+ tech keywords visualized
Kalkulator Cicilan — Installment calculator for Indonesian consumers
Alatwarung — Equipment catalog for street vendors