3. Data, Science & AI

Provides production-ready architectural patterns for LLM applications including RAG, agent systems, and observability.

Provides persistent memory systems for LLM conversations, supporting short-term, long-term, and entity-based memory retention.

Enables access to external AI models (e.g., DALL-E, GPT) for capabilities Claude lacks, including image generation and real-time data.

Expert in Langfuse, an open-source platform for LLM observability, enabling tracing, prompt management, evaluation, and monitoring of production LLM applications.

Optimizes LLM context windows using summarization, trimming, and routing to manage token limits and prevent context degradation in extended interactions.

Provides expert implementation patterns for Segment CDP, covering tracking, identity resolution, and data governance best practices.

Architects agent memory systems with short-term context management and long-term vector store retrieval strategies.

Creates custom, interactive data visualizations using d3.js for complex charts, graphs, and geographic maps in any JavaScript environment.

Provides best practices for building robust AI-powered products with LLM integration, RAG architecture, scalable prompt engineering, and cost optimization.

Provides optimized query patterns, analytics techniques, and data engineering best practices for high-performance ClickHouse database workloads.

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

Evaluates LLM agents via behavioral testing, capability assessment, and reliability metrics for real-world performance.

Validates TUI output using visual (PNG/SVG) and semantic (LLM) validation modes.

EvalKit enables conversational creation, execution, and analysis of AI agent evaluations via the Strands Evals SDK.

Analyzes social media campaign metrics including engagement, ROI, and audience insights to enable data-driven marketing strategies.

Optimizes Langfuse tracing performance to reduce latency and overhead in high-throughput LLM applications for production environments.

Installs and configures Mistral AI SDK/CLI authentication for seamless integration with Mistral's AI models and APIs.

Enables tracing of LLM calls and spans for observability in AI applications, facilitating debugging and performance monitoring.

Provides best-practice reference architecture and project layout for Databricks data engineering applications.

Provides best practices and patterns for implementing Langfuse SDK to track and analyze LLM application behavior.

Diagnoses and resolves common errors in Langfuse, an LLM observability platform, for debugging missing traces and integration issues.

Provides a minimal working example for integrating Langfuse tracing into LLM applications to track and debug model interactions.

Optimizes Mistral AI API performance using caching, batching, and latency reduction techniques for faster response times and higher throughput.

Automates GDPR-compliant data management in Delta Lake, handling PII, retention policies, and user data deletion requests.