3. Data, Science & AI

Optimizes Claude 4 (Sonnet 4.5) model responses via expert prompt crafting, chain-of-thought implementation, and clarity enhancement.

Migrates data from Neo4j Memory to Graphiti temporal knowledge graphs, ensuring data integrity and temporal consistency.

Enhances vague AI prompts through systematic research, clarification, and question generation to ensure effective execution.

Runs and monitors ATFT-GAT-FAN model training, hyper-parameter sweeps, and safety modes on A100 GPU infrastructure.

API for setting up Modal GPU sandboxes for interpretability research, used in setup.py with Sandbox, SandboxConfig, and ModelConfig.

Conducts comprehensive Meta Ads account audits, analyzing performance data, tracking infrastructure, and campaign metrics to generate detailed reports and architecture diagrams.

Configures and operates TinyDB for JSON-based document storage, enabling efficient data persistence and query management in applications.

Manages ElevenLabs conversational AI agents for initiating calls, retrieving transcripts, managing phone numbers, and analyzing conversations in voice AI development.

Generates Sora videos from text prompts using Azure OpenAI and saves the output as a local .mp4 file.

Extracts clean, ad-free article content from web URLs and saves as plain text for easy reading and reference.

Transforms Gmail into a business intelligence system using AI to analyze emails, detect leads, draft responses, and track client communications for MTL Craft Cocktails.

Standardizes vehicle insurance data by cleaning missing values, fixing formats, removing duplicates, and normalizing fields in CSV/Excel files.

Validates data quality in CSV/Excel files by checking required fields, formats, missing values, and generating quality reports for vehicle insurance platforms.

Processes vehicle insurance data from Excel/CSV files using Pandas, including cleaning, merging, validation, and business rule implementation.

Manages staff-institution mappings for vehicle insurance platforms via Excel-to-JSON conversion, conflict resolution, and coverage validation.

Manages data governance strategy, quality validation rules, and data dictionary for vehicle insurance platforms, including standards, metrics, and conflict resolution.

Analyzes vehicle insurance data from Excel/CSV files to generate business reports, monitor policy trends, and detect performance anomalies.

Analyzes Excel spreadsheets to summarize trends and recommend optimal chart types for data visualization.

Manages AI project artifacts in token-optimized YAML+Markdown format for efficient tracking, validation, and migration in AI development workflows.

Expertly crafts and optimizes prompts for Claude 4 (Sonnet 4.5) to enhance AI response quality, clarity, and effectiveness using Claude-specific techniques.

Sets up Python-Clojure interoperability for data science workflows, enabling NumPy, Pandas, and HuggingFace integration with data type conversion and performance optimization.

Provides a capability-based taxonomy for AI agent systems (Types A-D) with cost/latency analysis and multi-agent workflow optimization for architecture design.

Explains Google ADK architecture, agent patterns, and deployment for building production AI agents. Model-agnostic framework for designing and deploying AI systems.

Comprehensive overview of Clojure, Google ADK, and Vertex AI for building production AI agents, featuring a healthcare pipeline example with performance metrics.