3. Data, Science & AI

Transcribes Xiaoyuzhou podcast episodes into SRT subtitles and TXT transcripts using faster-whisper AI model.

Converts PDF documents to Markdown format, extracting text and images while fixing typographical issues for text analysis and content processing.

Performs complex AI model analysis and strategic planning using Cipher MCP, design pattern memory search, and multi-tool integration for LoRAIro.

Manages persistent storage and retrieval of contextual information, including memory file structures and update strategies for efficient context recall.

Enables end-to-end implementation of machine learning pipelines including model deployment, feature engineering, and monitoring for TensorFlow and PyTorch models.

Provides computational tools for statistical inference and optimization in data analysis and machine learning workflows.

Audits RAG systems for semantic collapse, precision drops, and architecture flaws in hierarchical and graph-based knowledge retrieval.

Architects and implements comprehensive simulation studies for scientific modeling and data-driven analysis.

Provides numerical algorithms and computational techniques specifically for statistical analysis and data processing tasks.

Designs and documents statistical algorithms with pseudocode and complexity analysis for data science applications.

Applies Directed Acyclic Graphs and potential outcomes frameworks to identify causal mediation effects in statistical analysis.

Provides essential mathematical frameworks and theoretical foundations for statistical analysis and data science applications.

Provides a structured methodology for constructing and verifying mathematical proofs in statistical research, enhancing rigor and reproducibility.

Conducts sensitivity analysis and assumption testing to assess model robustness and input dependencies.

Applies asymptotic statistical theory to causal inference, featuring M-estimation and semiparametric efficiency.

Offers a Python library for accessing, downloading, and analyzing French open data from data.gouv.fr with code examples and dataset documentation.

Enables querying, analysis, and reporting of machine learning experiment data and LLM traces from Weights & Biases.

Analyzes personal health data, medical records, lab results, and supplement protocols for health optimization and clinical insights.

Research pipeline for topology-aware GNN representation learning on power grids, enabling physics-guided models for power flow, OPF, and cascade prediction.

Creates domain expert systems with expertise files and workflows for integrating new technologies into AI platforms.

Intelligently designs and optimizes voice commands with context-aware execution, leveraging speech recognition for seamless user interactions.

Tracks recent IPO data and automatically generates TradingView watchlists for real-time market analysis.

Processes and generates multimedia content (audio, images, video, documents) using Google Gemini API for advanced vision, speech, and text-to-media capabilities.

Creates, refines, and optimizes prompts for LLMs using best practices to enhance model performance and reliability.