3. Data, Science & AI

Provides AI-driven second opinions on technical problems using GPT-5.2 via Codex CLI.

Executes SQL queries on PostgreSQL databases using MCP tools for efficient data retrieval and database operations.

Assists in writing and reading raw SQL queries for database interactions and data manipulation.

Manages LLM context windows by optimizing, compacting, and clearing context to prevent overflow and maintain model efficiency.

Step-by-step guide for defining and implementing custom agent types in the Beluga AI framework.

Generates visual summaries and actionable insights from Knack operational data for stakeholders through data processing and visualization.

Optimizes data queries for HTI dashboards and compliance reports through dynamic filtering and sorting, reducing API payload size and improving performance.

Integrates Leavn AI features including Stable Diffusion image generation, on-device models, and text generation for devotional, novelization, and kids content.

Provides REST API access to Knack objects and views for extracting and transforming operational data.

Integrates with Manus AI to delegate complex autonomous tasks including web browsing, code execution, and file generation.

Automates Knack API data retrieval by managing pagination to overcome 1,000-record limits, ensuring complete dataset access for inventory and compliance reporting.

Designs and builds AI agent workflows for Glif, enabling creation of new glifs, workflow structuring, and block type understanding.

Validates audio features to maintain coherence in Leavn's complex audio processing pipeline architecture.

Validates CloudKit data synchronization across multiple devices for Leavn's application.

Manages Bible features including verse search, translation handling, and AI transformations for scripture database operations.

Enables AI workflows in n8n using agents, LangChain, vector stores, RAG, and LLM providers for intelligent automation.

Specializes in parsing podcast RSS feeds to extract episode data, metadata, and media URLs for integration with Modcaster.

Specializes in on-device validation of audio machine learning models for Modcaster's AI audio processing system.

Automates data validation, deduplication, normalization, and integrity checks to ensure HTI compliance and dashboard accuracy.

Enables development and integration of AI agents, workflows, and AI-powered decentralized applications on the NEAR Protocol.

Runs Glif AI workflows through API for seamless integration of image and text generation capabilities into applications.

Specializes in audio fingerprinting and pattern detection for content analysis, enabling unique audio identification and matching in Modcaster's platform.

Accesses 150+ production-ready AI agent patterns for designing multi-agent workflows, feedback loops, and context management.

Configures OpenCode agents with permission settings, tool access, and behavior modes for secure and efficient AI agent deployment.