Jerome Etienne
I build AI tools that turn ideas into shipped products in days.
Projects
Can Machine Learning Predict Bitcoin?
Seven models, one honest answer
A 7-model ladder on Bitcoin 4-hour forecasting, from a zero-parameter naive baseline through ARIMA, XGBoost, LSTM, and Temporal Fusion Transformer, up to zero-shot and fine-tuned foundation models (Chronos-2, TimesFM). Same data, same test window, same scoreboard - read honestly, including the failures.
openai-cache
A caching layer for OpenAI API calls
A caching layer for OpenAI API calls, designed to improve performance and reduce costs. Wraps your existing OpenAI client to transparently return cached responses for repeated prompts, cutting both latency and your monthly bill.
openai-cost
Track the cost of each OpenAI call
Track the cost of every OpenAI API call your application makes. Get clear, per-request pricing visibility so you can monitor spending, identify expensive prompts, and keep your AI budget under control.
perplexity-cache
A caching layer for the Perplexity API
Cache for the Perplexity API. Avoid repeating the same search queries against Perplexity by serving identical requests from a local cache — faster responses, lower costs, and friendlier rate-limit usage.
video_transcript_whisper
Easily extract transcripts from videos
Extract transcripts from any video file with a single command. Powered by OpenAI's Whisper model, it turns audio into accurate, timestamped text — ready for search, subtitles, summaries, or downstream AI processing.
ContentMagick.ai
Produce your social media 10x faster
Our AI-powered tool simplifies content creation, generating engaging posts tailored to your audience. With enhanced formatting options, your messages will stand out on any social media platform. Say goodbye to writer's block and hello to creative, impactful content.
Contact Me
I would love to hear from you!