3. Data, Science & AI

Comprehensive toolkit for processing and analyzing physiological signals including EEG, ECG, and EDA for psychophysiology and biomedical research.

Performs differential gene expression analysis on RNA-seq data using DESeq2, including Wald tests, FDR correction, and volcano plots.

Simplifies RDKit operations for drug discovery tasks including SMILES processing, molecular descriptors, and 3D conformer generation.

Provides rapid bioinformatics queries via CLI and Python, accessing 20+ scientific databases including Ensembl, UniProt, and AlphaFold for sequence analysis and genomic data.

Analyzes database schemas, ORM entities, and performance issues across Java, Kotlin, and Python codebases, generating ERD documentation.

Optimizes prompts for context-aware AI responses and intelligently routes commands across multiple language models to maximize effectiveness.

Configures custom OpenCode agents with specialized prompts, tools, permissions, and models for LLM-based task delegation and agent management.

Creates, validates, and manages AI agent skills following the agentskills.io specification for structured agent capabilities.

Explains SpecWeave's context efficiency using Claude's progressive disclosure and sub-agent parallelization to minimize token usage and enable scalable AI operations.

Comprehensively evaluates machine learning models with metrics, cross-validation, and statistical significance testing, generating detailed visualized reports.

Manages ML experiment tracking with MLflow, Weights & Biases, and SpecWeave, ensuring reproducibility and integration with living documentation.

AI-driven quality assessment tool evaluating specifications, plans, and tests for clarity, testability, and risks, providing PASS/CONCERNS/FAIL decisions.

Expert in ksqlDB for real-time stream processing using SQL-like queries on Kafka topics, including windowing, joins, and materialized views.

Automates end-to-end feature engineering for ML pipelines, including data quality checks, feature creation, and production validation.

Provides ML-based anomaly detection using Isolation Forest, One-Class SVM, autoencoders, and statistical methods for fraud, intrusion, and system monitoring.

Generates human-readable model explanations using SHAP, LIME, and feature importance for trust, debugging, and compliance.

Manages Avro, Protobuf, and JSON Schema for Confluent Cloud and self-hosted environments, including evolution strategies and validation.

Builds end-to-end computer vision ML pipelines for image classification, object detection, and segmentation using PyTorch and TensorFlow.

Builds NLP pipelines for text classification, NER, sentiment analysis, and text generation using transformers and BERT.

Designs DAG-based MLOps pipelines for machine learning workflow orchestration using Airflow, Dagster, Kubeflow, or Prefect.

Automates hyperparameter tuning and model optimization for machine learning using Optuna, Hyperopt, and AutoML libraries with minimal manual effort.

Configures Kafka Connect for data pipelines, supporting source/sink connectors, CDC with Debezium, and SMT transformations.

Generates professional images for design assets using Pollinations.ai without API keys.

Provides time series forecasting using ARIMA, Prophet, LSTM, and statistical methods for trend decomposition, seasonality detection, and multivariate predictions with confidence intervals.