3. Data, Science & AI

Provides call statistics for the Vocal Bridge voice agent, including total calls, success rate, and average duration.

Verifies end-to-end LLM behavior using real OpenAI and Gemini APIs, ensuring actual model performance beyond stub tests.

Aggregates communications from all channels into a centralized, queryable event store for unified data access.

Guides the design, implementation, and maintenance of robust data models to ensure data integrity, performance, and scalability.

Provides a guide for interacting with PostgreSQL databases using the DBManager interface, covering key tables and common query patterns.

Verifies claims in AI-generated content against external sources to detect and correct hallucinations, used as a post-generation step.

Analyzes Excel spreadsheets to generate insights via pivot tables, charts, and tabular data processing for data-driven decisions.

Enables semantic search and analysis of AI conversation history across multiple development tools using embedding-based retrieval.

Provides a step-by-step checklist for integrating a new sport into a betting system, covering Elo ratings setup, game data ingestion, and workflow configuration.

Searches and retrieves past AI agent sessions, stored knowledge, and indexed documentation for efficient information recall and knowledge management.

Queries Google NotebookLM notebooks for source-grounded, citation-backed answers using Gemini, reducing hallucinations via document-only responses.

Orchestrates multiple AI agents with specialized expertise to handle independent tasks or provide comprehensive analysis from diverse perspectives.

Analyzes model susceptibility to reward hacking via prefill sensitivity, comparing logprob and token-count metrics for exploitability prediction.

Provides token-efficient tracking for AI orchestration workflows, enabling CLI-based status updates and agent-assisted complex operations.

Optimizes generative AI search engines (ChatGPT, Claude, Perplexity) for enhanced query relevance and performance.

Guides on database schema design, indexing strategies, ORM selection, and serverless database implementation.

Manages context networks for AI agents, ensuring effective information retrieval and agent performance throughout project lifecycles.

Analyzes conversation history to identify learnings and corrections, then proposes system updates to enhance future performance. Triggered by reflection commands.

Adapts AI behavior across task-specific modes including brainstorm, implement, debug, and review.

Manages AI-powered nutrition coaching with personalized advice, daily tips, motivational challenges, and gamified streak tracking.

Shows a sample output of the Research-to-Roadmap pipeline without executing analysis, demonstrating expected results.

Builds servers following Model Context Protocol (MCP) principles for efficient AI model context management and resource handling.

Calculates nutritional values (calories, proteins, carbs, fats) and portion estimates from food items for dietary analysis and meal planning.

Configures comprehensive safety settings for Gemini AI agent across all content categories to prevent harmful outputs.