3. Data, Science & AI

Builds end-to-end MLOps pipelines for data preparation, model training, validation, and production deployment automation.

Manages agent memory across hot, warm, and cold tiers with event consolidation and scheduled cleanup to optimize performance and storage after task completion.

Reference guide for LLM model IDs, capabilities, pricing, and provider endpoints to aid in model selection and validation for AI projects.

Monitors, manages, and debugs Dagster data pipelines for local and Kubernetes deployments.

Generates images, videos, and audio using Google's Gemini AI models, supporting text-to-image, image-to-image, text-to-video, and image-to-video workflows.

Converts meeting audio files (MP3, WAV, M4A) to text using AWS Transcribe's AI-powered speech recognition service.

Performs database operations (backup, update, search, query) for XStore and WPBakery WordPress themes using WP-CLI and SQL.

Assists in developing deep learning projects including model architecture, training, and deployment workflows.

Optimizes LLM performance and output quality through advanced prompt engineering patterns for production deployment.

Queries the v3_openapi database to retrieve integrated connector and API configuration details via SQL.

Transcribes audio files with speaker diarisation and automatic speaker name identification using Parakeet MLX, an AI model for speech processing.

Command-line interface for building, running, and orchestrating AI agents with Claude, supporting authentication, headless mode, and JSON output parsing.

Automates client-ready dashboard creation from data files with pattern detection, visualizations, and actionable insights for professional presentations.

Coordinates parallel execution of AI agents for complex multi-step tasks using background processing and delegation patterns.

Researches solutions, algorithms, and best practices using SurrealDB GraphRAG knowledge base with 600+ scientific papers, StackOverflow, and GitHub repositories.

Processes financial data including trade imports, position management, and risk metric calculation for trading systems.

Transforms Claude into a senior AI/ML engineer for building production-grade ML systems, LLMs, and computer vision solutions from scratch.

Counts words in JSONL text fields, supporting auto-discovery or specific field selection across multiple files for text analysis and comparison.

Analyzes JSONL benchmark data for errors, length issues, and text quality with pre-filtering capabilities for validation tasks.

Executes Fabric patterns natively within Claude for content processing tasks including summarization, claim analysis, and threat modeling.

Optimizes response sizes from MCP tools to prevent conversation overload using sophisticated token management for AI efficiency.

Provides best practices for integrating Google Gemini API, including prompt engineering, streaming responses, and multimodal processing for AI applications.

Generates automated PySpark tests to validate data completeness, accuracy, and consistency in Databricks tables.

Transforms unstructured human intent and repository data into dynamic, operational domain profiles using knowledge engineering techniques.