3. Data, Science & AI

Provides Bayesian parameter estimation and posterior sampling via emcee MCMC for physics data analysis.

Provides mathematical methods for function interpolation and approximation, essential in numerical analysis and data modeling.

Automates ACMG/AMP guideline-based classification of genetic variants to assess pathogenicity in clinical genomics workflows.

Calculates CMB power spectra, matter power spectra, and estimates cosmological parameters using the CAMB code for cosmological research.

Configures Microsoft AutoGen for multi-agent conversational AI systems, enabling complex dialogue management and agent collaboration.

Provides ELECTRE family methods for outranking-based decision support, analyzing concordance and discordance in multi-criteria decision problems.

Implements standard metadata schemas (Dublin Core, METS, MODS) for organizing and describing digital collections and archival materials.

Enables quantum circuit design and execution using Google Cirq on quantum processors for scientific and AI applications.

Enables particle physics data analysis using Scikit-HEP toolkit for high-energy physics datasets with modern Python tools.

Provides RAG system setup and management via the unrag TypeScript library, covering installation, ContextEngine API, and component configuration.

Processes YouTube videos to extract transcripts, generate infographics, and create audio summaries using AI and data processing techniques.

Enables implementation of code leveraging the Sonnet AI model with full main context access for seamless integration.

Guides users through a multi-AI workflow: plan, review, implement, and review, while cleaning old task files.

Guides software architects in evaluating and implementing goal-seeking agent design patterns for autonomous problem-solving in AI-driven systems.

Automates AI model benchmarking, comparison, and performance testing using Benchmark Suite V3.

Proactively manages AI model context windows through token monitoring, intelligent extraction, and selective rehydration to prevent token limit issues.

Manages AI model context windows by monitoring tokens, extracting key information, and selectively rehydrating to avoid token limits while preserving essential context.

Provides comprehensive knowledge for building production AI agents and workflows using Microsoft Agent Framework, including auto-activation for agent design and enterprise AI tasks.

Tracks and reports AI agent metrics including invocation counts, success rates, and completion times to optimize agent utilization and delegation patterns.

Analyzes Claude session trace JSONL files to identify health issues, token usage patterns, and failure causes for debugging.

Analyzes chemical processes using molecular structure, reaction mechanisms, and analytical techniques to evaluate reactions, materials, and environmental impact.

Applies physics principles and quantitative modeling to analyze energy systems, physical constraints, and technological feasibility.

Runs Microsoft's eval-recipes benchmarks to validate AI model improvements (amplihack) against baseline agents, auto-activating during testing and benchmarking.

Guides architects in evaluating and implementing goal-seeking agent design patterns for autonomous problem-solving in AI systems.