3. Data, Science & AI

Guides robust implementation of AI-powered products through LLM integration, RAG architecture, scalable prompt engineering, and cost-optimized AI UX design.

Provides persistent memory systems for LLM conversations with short-term, long-term, and entity-based context retention across interactions.

Assists in building and optimizing transformer models, covering self-attention, multi-head attention, RoPE, and positional encoding for AI and NLP applications.

Assists in designing, modeling, and optimizing renewable energy systems including solar PV, wind power, and grid integration.

Generates data-driven charts and visualizations using Python, supporting multiple chart types for business and research analysis.

Enables local browser AI execution using Transformers.js, WebLLM, and ONNX Runtime for privacy and zero API costs.

Enables implementation of RL algorithms including PPO, Q-learning, and RLHF for training AI agents and aligning LLMs with human feedback.

Provides patterns for building AI applications that integrate text, images, audio, and video using vision APIs, audio transcription, and unified pipelines.

Orchestrates parallel execution across multiple AI models (Claude, GPT, Gemini, Grok) and synthesizes the best answer from their outputs.

Specializes in designing and maintaining data pipelines for ETL, CDC, batch, and stream processing with emphasis on data quality and transformation.

Provides community health metrics, engagement analytics, sentiment analysis, and cohort tracking to drive data-informed decisions for community management.

Provides tools for constructing optimized investment portfolios using modern portfolio theory techniques including mean-variance optimization and risk parity.

Provides architecture guidance for agent memory systems including short-term, long-term, and cognitive frameworks (semantic, episodic, procedural) with emphasis on retrieval strategies and vector store integration.

Expertly designs and evaluates prompts for LLM applications, optimizing structure, context, and output formatting to enhance model performance.

Provides tools for building quantitative risk models including VaR, stress testing, and Monte Carlo simulations across market, credit, and operational risk frameworks.

Expert in LangGraph, a production-grade framework for building stateful, multi-actor AI applications with agent workflows, state management, and ReAct patterns.

Architects autonomous AI agents using LLM frameworks, emphasizing tool integration, memory systems, and multi-agent orchestration.

Applies scientific method to data science and computational research, emphasizing hypothesis testing, experimental design, and reproducibility while avoiding statistical pitfalls.

Provides patterns for physics-based simulations including numerical integration, rigid body dynamics, and fluid simulation, emphasizing accuracy, stability, and performance.

Manages LLM context windows through summarization, trimming, and routing to prevent overflow and maintain model effectiveness.

Specialist in PostgreSQL query optimization, indexing, and advanced database features.

Provides comprehensive observability for LLM applications with tracing, cost tracking, RAG evaluation, and hallucination detection to optimize AI performance and output quality.

Expertly builds Retrieval-Augmented Generation systems with optimized embeddings, vector databases, and retrieval strategies for LLM applications.

Optimizes AI model interactions for creative outputs through structured prompting, iterative refinement, and model-specific strategies.