3. Data, Science & AI

Expert in PostgreSQL and Supabase database management: schema design, RLS policies, query optimization, and Supabase-specific features.

Retrieves task-relevant contextual memories from a memento knowledge graph at initialization to enhance AI decision-making.

Builds LLM-powered applications with LangChain, featuring RAG pipelines, agent frameworks, and memory management for AI integration.

Computes dsRNA candidates using a sliding window algorithm for bioinformatics applications.

Executes a complete double-stranded RNA (dsRNA) design workflow with mandatory user confirmation at each step for scientific accuracy.

Builds AI agents, RAG pipelines, and LLM-powered applications using LangChain in Python or TypeScript.

Analyzes Ahrefs site audit CSV data to identify SEO issues, generate summaries, and create actionable fix checklists for website health.

Builds AI agents for code execution, tool calling, multi-agent systems, and agentic RAG using Hugging Face's SmolAgents framework.

Executes standardized clinical formulas for medical trial procedures, ensuring accurate calculations.

Identifies essential genes in a genome by leveraging orthologous gene relationships and scientific literature.

Fetches coding DNA sequences (CDS) from NCBI RefSeq for a specified species.

Generates a comprehensive markdown report with ranked candidate data for analysis and review.

Guides users through Q&A to design, configure, and optimize AI agents for software development workflows.

Compiles multiple run findings into a single report with cross-run diffs and stable conclusions.

Performs backtesting and metrics analysis for trading strategy evaluation.

Builds stateful agent graphs with conditional flows, parallel execution, and multi-agent orchestration for AI applications.

Analyzes CSV files to generate statistical summaries and visual plots using Python's pandas library.

Screens dsRNA candidates for off-target matches in human and honeybee genomes using bioinformatics tools to ensure target specificity.

Calculates composite scores by integrating efficacy and safety metrics for evaluation and ranking purposes.

Automates review and filtering of JorBox game cards from CSV data to ensure cultural appropriateness, Halal compliance, and quality standards.

Generates and edits images via Gemini AI model, supporting text-to-image, style transfers, and multi-reference composition for creative assets.

Extracts audio from dashcam MP4 files and generates GPU-accelerated, timestamped transcripts with speaker diarization for conversation analysis.

Extracts 3 key frames (BEGIN, MIDDLE, END) from Movie_F dashcam videos using GPU-accelerated processing and gap analysis.

Optimizes system prompts for Claude Code agents using proven prompt engineering techniques to improve agent performance and behavior in workflows and tool instructions.