Data & AI skills

Structure PyTorch training code with PyTorch Lightning, separating model logic, data preparation, and training settings. Use templates and configuration examples to add checkpoints, experiment logs, and distributed training that runs across multiple GPUs or TPUs.

• Restructuring PyTorch code into LightningModules
• Building reusable data preparation pipelines
• Configuring training across multiple GPUs or TPUs

Create scientific plots for manuscripts with consistent layouts, accessible colors, and error bars showing uncertainty. Use matplotlib, seaborn, or plotly to prepare multi-panel figures and export files with publication-specific sizes and resolution.

• Prepare figures for journal submission
• Arrange multiple plots with consistent styling
• Improve color and grayscale readability

Plan experiments, analyze data, and build and evaluate predictive models. The skill also covers explaining findings to stakeholders and designing production workflows for model deployment, monitoring, and retraining to support business decisions.

• Designing experiments and A/B tests
• Preparing model inputs and evaluating predictions
• Planning model deployment and monitoring

Build image and video AI systems for object detection and segmentation, which identifies separate image regions. The skill covers custom model training, dataset pipelines, inference optimization, real-time processing, and deployment and monitoring in production.

• Developing object detection and image segmentation systems
• Training custom vision models with project data
• Optimizing inference for real-time video processing

Build and evaluate machine learning models in Python with scikit-learn. Covers classification, numerical prediction, grouping similar data, preprocessing, model comparison, and pipelines that combine data preparation and prediction into a repeatable workflow.

• Build classification or numerical prediction models
• Group customers with similar characteristics
• Compare models using cross-validation

Use RDKit in Python to read molecular structures, calculate properties such as molecular weight, find similar compounds or specific structural features, and generate molecular drawings and 3D coordinates for chemistry research.

• Calculating properties for a list of molecules
• Finding compounds similar to a reference molecule
• Filtering molecules by a specific structural feature

Explain machine learning predictions with SHAP values, which measure how each input contributes to an output. Create charts for individual predictions and overall feature importance, investigate errors, compare models, and examine differences between groups.

• Explaining an individual prediction
• Visualizing overall feature importance
• Investigating misclassifications and data leakage

Build searches that find documents by meaning and support recommendation features using vector databases. Includes implementation patterns for storing and retrieving vectors, combining meaning with keyword matches, and tuning search speed and result quality.

• Build document search based on meaning
• Implement retrieval for RAG, AI answers grounded in retrieved documents
• Combine keyword and meaning-based search

Guides single-cell RNA-seq analysis, which examines gene activity in individual cells. Covers data loading, quality checks, clustering, marker genes, cell type annotation, and plots, with AnnData handling and scVI/scANVI workflow planning.

• Checking and filtering single-cell RNA-seq data
• Visualizing cell groups with UMAP or t-SNE
• Finding marker genes and annotating cell types

Build production systems for machine learning models, from deployment to monitoring and retraining. The skill also guides LLM integration and retrieval augmented generation, which uses retrieved information to support answers, with scripts and architecture references.

• Deploying machine learning models to production
• Setting up model monitoring and retraining
• Building applications that answer using retrieved information

Process mass spectrometry data in Python with PyOpenMS. Read common file formats, smooth and normalize signals, detect peaks, align samples, and process peptide identification results for protein and metabolite quantification workflows.

• Reading mzML files and inspecting spectra
• Smoothing and normalizing mass spectrometry signals
• Aligning and comparing metabolite peaks across samples

Analyze how long events take to occur, including records where observation ends before the event. Use Python to prepare data, fit and compare survival models, and evaluate predictions with measures suited to incomplete observations.

• Analyze time to events with incomplete observations
• Compare Cox models and Random Survival Forests
• Select features in datasets with many variables

Build Python simulations with SimPy to study queues and shared resources such as machines, staff, or servers. Model events over time, measure waiting times and utilization, and compare capacity choices.

• Compare service waiting times under different staffing levels
• Measure machine utilization on a production line
• Model inventory flows between producers and consumers

Plan and audit preprocessing for grouped time-series data, such as matches or patient records. Check missing-value handling, scaling scope, feature meaning, and data leakage, where information unavailable at prediction time enters model inputs.

• Planning missing-value handling within each group
• Choosing within-group or global scaling
• Auditing preprocessing code for data leakage

Create statistical charts in Python with seaborn to explore distributions, compare groups, and examine relationships between variables. Build scatter plots, heatmaps, regression charts, and multi-panel figures, then adjust styling and export figures for reports or publications.

• Exploring data distributions and outliers
• Comparing measurements across groups
• Visualizing correlations with a heatmap

Analyze biological data in Python, from DNA, RNA, and protein sequences to evolutionary trees and microbial communities. Read and convert files, compare diversity within and between samples, and test differences between groups.

• Processing DNA, RNA, and protein sequences
• Building and comparing evolutionary trees
• Comparing microbial community diversity

Design and improve prompts and structured responses for AI applications. The skill also guides evaluation, retrieval augmented generation, which uses retrieved material to inform answers, and agent architecture, with references and scripts for production workflows.

• Improving prompts and response formats
• Evaluating answers based on retrieved documents
• Designing AI agent systems

Use SAELens to train and analyze sparse autoencoders, which break model activations into a small set of interpretable features. Explore pretrained features, evaluate reconstruction quality, troubleshoot training, and test how individual features influence generated text.

• Exploring features in a pretrained SAE
• Training an SAE for a selected model layer
• Investigating inactive features and reconstruction errors

Apply NOWAIT to shorten reasoning model outputs by suppressing reflection words during generation. Includes integration examples for HuggingFace Transformers and vLLM, with guidance on model suitability and potential accuracy loss in distilled models.

• Reducing tokens used in lengthy reasoning
• Introducing NOWAIT into model inference
• Checking model suitability and accuracy risks

Estimate intervention effects from existing data using causal analysis methods. Set up models, compare observed outcomes with estimates of what would have happened without intervention, check robustness, and interpret fitted results.

• Estimating intervention effects from existing data
• Comparing outcomes with and without intervention
• Checking the robustness of causal estimates

Create interactive charts in Python for exploring scientific, statistical, geographic, and financial data. Choose quick plotting or detailed customization, add hover details and zoom controls, and export results as interactive HTML or static images.

• Exploring measurement relationships with scatter plots
• Comparing distributions across groups
• Inspecting time series with a range slider

Work with microscopy images stored in OMERO through its Python API. Retrieve datasets, analyze pixel values and selected image regions, manage annotations and metadata, and save measurements or automate processing across multiple images.

• Retrieving microscopy images from OMERO
• Measuring pixel intensity in selected regions
• Saving image tags and analysis results

Find and retrieve AI prompt templates through prompts.chat, filtering by keywords, categories, tags, or media type. Improve existing prompts for text, image, video, or sound tasks, with an explanation of the changes.

• Find prompt templates for code reviews
• Browse prompts for writing tasks
• Retrieve a specific prompt by ID

Develop healthcare prediction models from electronic health records and physiological signals. Use PyHealth to prepare data, train models for tasks such as mortality or readmission prediction, and evaluate accuracy, fairness, and how predictions can be explained.

• Building mortality prediction models with MIMIC-IV
• Training and evaluating hospital readmission models
• Developing medication recommendation models with SafeDrug