3. Data, Science & AI

Provides foundational guidance for running the General Lake Model (GLM) to simulate lake temperature dynamics, including input file interpretation and configuration adjustments.

Provides fuzzy string matching capabilities to reconcile entity names (e.g., companies, people) across datasets with spelling variations, typos, or formatting differences.

Provides a Python package for the GaMMA earthquake phase association algorithm, processing seismic phase picks and station data to cluster events and output source parameters.

Summarizes advantages and disadvantages of seismic phase picking methods based on 2025 Earthquake Catalog Workshop research.

Comprehensive tool for spreadsheet creation, editing, and data analysis with formula support, formatting, and visualization capabilities.

Processes GLM model output files, extracting temperature data, converting depth coordinates, and calculating RMSE against observational data.

Analyzes STL mesh files to compute geometric properties (volume, components) and filter debris from noisy 3D scan data.

Analyzes images and multi-frame sequences using OpenAI GPT vision models for computer vision tasks.

Automates cleaning of messy tabular data including deduplication, missing value handling, outlier removal, and text normalization.

Builds scalable data pipelines, ETL/ELT systems, and real-time streaming infrastructure using Spark, Airflow, and Kafka for data modeling and governance.

Provides client API access to download earthquake data from FDSN data centers using ObsPy, enabling seismological research data acquisition.

Provides spatial utilities for hexagonal grid calculations including distance, neighbor finding, and spatial queries in offset coordinate systems.

Parses power system network data (bus, generator, branch) for power flow analysis in electrical engineering.

Converts SQLite databases and map data into structured JSON for schema exploration and data extraction.

Provides an overview of ObsPy's core data API for parsing seismological file formats and standardizing data for signal processing and modeling workflows.

Detects anomalies in time series data using Meta's Prophet framework to model trends, seasonality, and holiday effects.

Extracts text content from images using Tesseract OCR via Python, enabling automated data capture from visual sources.

Toolkit for matching entity names across datasets with spelling variations, typos, or formatting differences, enabling accurate data reconciliation.

Detects anomalies in time series data using Meta's Prophet framework, identifying unusual trends and patterns for data analysts.

AI-powered video analysis via Google Gemini API for summaries, Q&A, transcription with timestamps, scene detection, and multi-video comparison.

Efficiently trains GPT-2 scale language models (124M parameters) on a single GPU with mixed precision and modern optimizers.

Performs multi-hop evidence search and structured extraction across enterprise documents, chats, meetings, and PRs, returning JSON-ready entities with evidence pointers.

Detects speech segments in audio files using VAD tools for preprocessing in speech analysis and speaker diarization.

Implements Manifold-Constrained Hyper-Connections (mHC) algorithm for deep neural network training stabilization using Sinkhorn-Knopp-based doubly stochastic matrices.