3. Data, Science & AI

Translates business goals into data metrics and KPI dashboards for performance tracking.

AI-powered code and plan review using Gemini 3 Pro for large context processing (over 200k tokens).

Builds MCP servers for LLMs to interact with external APIs, supporting Python (FastMCP) and Node/TypeScript (MCP SDK).

Provides a production-ready framework for building and orchestrating multi-agent AI systems with enterprise features, including RAG, vision, and multimodal agents.

An AI knowledge package manager enabling on-demand discovery, evaluation, and installation of skills from local or global repositories to extend agent capabilities.

Enables querying, configuring, and optimizing QLever, an open-source RDF triplestore, with support for SPARQL, GeoSPARQL, and text search operations.

Manages long context inputs for LLMs using MIT CSAIL's Recursive Language Model approach, enabling efficient processing of large codebases and multi-file operations.

Generates professionally styled Jupyter notebooks using markdown, CSS, and HTML, adhering to Airbnb Design Language System for data science documentation.

Framework for building AI agents, LLM integration, and tool calling workflows using Kotlin.

Designs and manages semantic knowledge systems using SKOS, including taxonomies, thesauri, and AI/LLM integration patterns for structured data representation.

Dispatches multiple AI agents in parallel to investigate and resolve independent failures without shared state.

Provides context-efficient, read-only delegation to specialized subagents within AI systems to optimize task handling.

Designs and implements OWL 2 ontologies for structured knowledge representation, covering RDFS, profiles, and reasoning patterns for AI data integration.

Writes, analyzes, and optimizes SPARQL 1.1 queries for knowledge graphs using RDF, OWL, SHACL, and Turtle serialization with AI integration.

Provides real-time access to Plaid's banking API for account, transaction, and balance data retrieval.

Generates periodic reports featuring metrics, interactive dashboards, and data analysis for business insights.

Converts complex PDFs to structured Markdown with preserved hierarchy, tables, images, and layout using AI layout analysis.

Validates RDF data against SHACL constraints, ensuring semantic data integrity and compliance with W3C standards. Integrates AI/LLM patterns and taxonomy systems.

Researches and optimizes ComfyUI workflows, LoRAs, and custom nodes for AI image generation using community resources and API-driven manipulation.

Analyzes and reconstructs user prompts using proven frameworks to eliminate ambiguity and maximize AI response quality.

Generates structured prompts for autonomous AI agents, creating PROMPT.md with phases, validation criteria, and completion promises from PRD or rough descriptions.

Delegates complex reasoning, planning, and deep analysis tasks to OpenAI's GPT-5.2-Codex model for extended thinking and multi-step problem-solving.

Launches parallel AI agents to process labeled tasks using a template and confirms execution.

Provides best practices for ClickHouse database patterns, query optimization, and analytics workflows to enhance high-performance data processing.