3. Data, Science & AI

Optimizes AI model performance through structured few-shot example patterns, ensuring consistent and high-quality outputs beyond zero-shot capabilities.

Provides a structured workflow for designing, implementing, and verifying data transformation pipelines with schema mapping and ETL process support.

Specializes in designing system prompts and role prompting techniques to optimize AI agent behavior and generate high-quality responses.

Provides expert guidance on SQLite index design strategies including B-Tree, partial, expression, and covering indexes for optimal query performance in Drizzle ORM environments.

Diagnoses and optimizes slow SQLite queries using EXPLAIN QUERY PLAN, index strategies, and query rewriting.

Prevents AI hallucinations using three-layer defense (prompt, parameter, verification) to guarantee factual, accurate, and verifiable outputs.

Optimizes AI/ML command performance by enhancing token efficiency, parallel execution, and model selection to reduce latency and resource consumption.

Optimizes database query performance by resolving N+1 problems, analyzing execution plans, and designing efficient indexes and fetch strategies.

Executes intent validation and synchronization checks by triggering dedicated agents for full or targeted operations.

Identifies and resolves SQL anti-patterns through schema reviews, query analysis, and performance optimization for improved database design and efficiency.

Manages end-to-end AI agent lifecycle operations including initialization, state handling, versioning, and maintenance for continuous quality assurance.

Designs structured role prompts for AI agents, emphasizing responsibility separation, expert thinking styles, and software design principles for effective system interactions.

Designs foreign key constraints using C.J. Date's theory, implements CASCADE strategies, resolves circular references, and ensures soft delete consistency for Drizzle ORM.

Optimizes AI model context by minimizing token usage and efficiently extracting necessary information via lazy loading and index-driven design.

Designs prompts to select reasoning patterns, compare solution paths, and generate consistent, evidence-based conclusions for AI reasoning tasks.

OpenProse structures natural language into unambiguous control flow for AI sessions, enabling multi-agent workflow orchestration and AI task execution.

Specializes in testing, evaluating, and improving AI prompts through A/B testing, automated metrics, and iterative quality assurance for reliable LLM performance.

Generates domain-specific reports by processing and visualizing data, activated on demand for data-driven insights.

Manages domain-specific level systems, activating on demand to handle tiered task execution and progression.

Aggregates onchain data from Talent Protocol, Guild.xyz, Basenames, Zora, and DeBank to enrich creator profiles and reputation scores for discovery and ranking.

Creates new OpenCode AI agents with GPT-5.2-Codex as the default model configuration.

Accesses ATXP API for real-time web search, AI-generated images, music, video, and X/Twitter search with authentication.

Provides tools for creating, debugging, and deploying Concordance mods to customize LLM behavior during inference, supporting events and actions like force_tokens and SelfPrompt.

Expert in implementing and managing geospatial data catalogs using the STAC specification for efficient data discovery and access.