3. Data, Science & AI

Provides end-to-end data engineering solutions for ML systems including pipelines, quality assurance, storage, and versioning.

Conducts end-to-end AI system evaluation including model selection, benchmarking, cost-latency analysis, and build-vs-buy decisions.

Enables building RAG pipelines with chunking, embeddings, vector search, and query transformation for knowledge-augmented applications.

Provides insights into foundation model architecture, configuration parameters, and behavior for effective LLM deployment and usage.

Optimizes AI models for production through quantization, pruning, and hardware-aware tuning to reduce resource consumption without compromising performance.

Provides expert review and design assistance for Algebraic Intermediate Representations (AIR) and zero-knowledge constraint systems in STARK protocols.

Manages training data lifecycle including labeling, augmentation, imbalance handling, and splitting for optimal model performance.

Provides comprehensive knowledge for building production-ready machine learning systems from data engineering through deployment and operations, based on Harvard's ML Systems course and Chip Huyen's book.

Enables building reliable production ML systems with model monitoring, drift detection, and failure handling capabilities.

Efficiently extracts structured data from websites and APIs using programmatic methods, avoiding context overload.

Configures DigitalOcean Gradient AI for serverless LLM inference, model access keys, and AI agent development using Agent Development Kit (ADK).

Provides a primer on deep learning fundamentals including neural networks, backpropagation, optimization, regularization, and training best practices.

Enables AI-powered web search, content extraction, and site structure mapping via Tavily's API for data gathering.

Executes remote experiments and processes results using the doe-suite framework, generating visualizations like plots for data analysis.

Coordinates multiple AI models (ChatGPT, Gemini, etc.) to verify specification-code-test consistency, filter false positives, and reduce manual review time.

Generates PhD-level domain-specific AI agents for niche expertise where standard solutions are inadequate.

Enforces best practices for TimescaleDB 2.24.0, optimizing time-series data handling with hypertables, continuous aggregates, and compression.

Guides step-by-step implementation of research papers into AI/ML models from scratch, using checkpoint questions to build understanding and ensure accurate reproduction.

Deploys ML models on Hugging Face Spaces using Gradio, Streamlit, and Docker for chat, image generation, and model comparisons.

Automates Sentry configuration for AI agent and LLM monitoring across SDKs like OpenAI, LangChain, and Anthropic.

Enables AI systems to recall past conversations and persistently store facts using episodic memory and knowledge graphs.

Enforces PostGIS 3.6.1 best practices for spatial queries, geometry operations, and location-based data processing.

Designs specialized AI agents with reusable prompts, triggering conditions, and validation tools for efficient task automation.

Builds conversational AI agents with external tool access using OpenAI Agents SDK and Model Context Protocol (MCP) across multiple AI providers.