Expert AI data analyst that analyzes any CSV, Excel or tabular dataset. Paste or upload data and get: 📊 Column profiling & statistics · 💡 Key patterns & anomalies · ⚠️ Anomaly detection · ✅ Data quality score · 🔗 Correlation insights · 🚀 Actionable recommendations. Works with sales, financial, survey, log, and any structured data. Educational purposes only.
Four-layer AI detection evasion, not word swapping. Real math (burstiness CV, vocabulary density) — no guessing. Detector-specific: attacks GPTZero, Originality, Turnitin by their weakest links. Chinese NLP: tracks 四字成语密度, a blue ocean. 105+ AI markers with genre-aware replacements. Research-backed (27 reports, 25+ papers). Honest: 75-85% typical improvement, no fake 100% claims.
Spot what looks unusual in Binance USDⓈ-M perpetuals: funding-rate percentiles, open interest vs price divergence, basis deviation and long/short ratio jumps — each with its comparison baseline and a data-quality grade. Read-only public data. No trading, no account, no API key.
Find the data problems that silently hurt your model before you train: suspected mislabels, duplicate and near-duplicate images, train/val/test leakage, conflicting labels, class imbalance, corrupt images, missing label files and broken bounding boxes. Supports classification folders, YOLO and COCO. Runs locally and read-only, so your data never leaves your machine. Get a prioritized fix list in plain language.
Put in one video and SceneBook finds cut boundaries with ffmpeg scene detection (threshold sweep + merging short scenes), pulls a key frame per scene, and builds an A4 landscape contact sheet PNG/PDF, splitting pages when full. A review version with a blank comment box per card, a scene list (CSV, Markdown) and a cut-length bar chart with an average line come too. On a synthetic sample (10 scenes, known boundaries) it found all 10 with 0 ms boundary error. Local, no API key.
Audits soccer analysis before conclusions are trusted: source tiers, dates and seasons, competition splits, 450-minute sample checks, availability flags, and uncertainty labels.
Analysez logs, diagnostics, erreurs et éléments techniques fournis afin d’identifier des causes probables, prioriser les anomalies et préparer un plan de correction.
A full analysis pipeline that turns a dataset into a finished report, not just charts. Hand it a CSV, Excel, Parquet, or JSON file and it cleans the data, scores its quality, runs EDA, and where the data supports it, trains a prediction model with SHAP attribution and a backtested forecast. You get a trust score, ranked findings, and recommendations as a Word doc, an interactive HTML report, and an optional PowerPoint deck. The numbers come from code, so they stay consistent, never invented.
Paste AI evaluation cases and get an evidence-bounded triage brief: expected vs. observed behavior, ranked hypotheses, and the missing checks to run next — so you can decide what to investigate first without claiming a root cause or fix.