3. Data, Science & AI

Creates and manages reusable templates for consistent AI prompt generation, optimizing interactions with language models.

Automates transcription and insight extraction from walk recordings, converting audio data into structured research material for pipeline integration.

Creates reusable, structured metaprompts for consistent AI task execution across various AI tools and workflows.

Enables compact Marimo notebook patterns with reactive cells and interactive data visualizations for streamlined data exploration.

Intelligent context retrieval and learning agent that injects strategies before answers and stores insights after tasks to enhance AI reasoning.

Provides persistent memory for Claude AI, automatically compressing context to reduce token usage by 84% and save API costs while managing project context across sessions.

Extracts entities and concepts from documents using AI memory systems for knowledge base population and content structure analysis.

Extracts and processes repository data from GitHub for analysis, integration, or data pipeline purposes.

Creates intent-based routing components to distribute tasks among AI agents according to user intent.

Compares and evaluates embedding models to determine optimal performance for semantic search tasks.

Enables semantic knowledge retrieval and pattern analysis from stored data using Qdrant vector database for efficient research.

Aggregates results from parallel AI agents with distinct prompts to identify gaps, patterns, and friction in complex problem spaces.

Offers step-by-step guidance for incorporating new data sources into RFP workflows beyond the RFPMart platform.

Exports RFP, evaluations, and pursuits data in CSV and other formats for analysis and reporting purposes.

Provides compact Snakemake workflow patterns by keeping rules concise and moving complex logic to Python modules for efficient data analysis pipelines.

Provides optimized patterns for ingesting knowledge into vector databases to enhance RAG system performance.

Provides schema verification for haios_memory.db to confirm table and column names before executing SQL queries or database operations.

Provides a defensive epistemology protocol for coding agents, enforcing explicit predictions and verification to ensure reliable code generation.

Enables document storage and retrieval via Qdrant for RAG workflows, supporting semantic search and persistent memory systems.

Applies Recursive Language Model (RLM) reasoning to analyze large files, codebases, and documents, enabling efficient processing of contexts exceeding 1000 lines.

Extracts structured data from websites through web crawling and content parsing.

Manages persistent memory and context compression for Claude Code, optimizing token usage and enabling efficient insight recall.

Provides systematic patterns for data analysis, pattern recognition, and troubleshooting to enhance data-driven decision making.

Provides patterns to integrate RAG context from Qdrant into any agent, enabling persistent memory for external or imported agents.