Daniel Podrażka - AI & Data Engineer

About Me

I'm Daniel Podrażka, an AI & Data Engineer with expertise in building AI-powered systems, trading infrastructure, and LLM-based applications. I help businesses optimize their AI systems and make strategic technology decisions that drive real business value through intelligent automation and data-driven insights.

AI Architecture Assessment

Evaluating and optimizing AI system architectures, recommending the right technology stack for your specific use case and business requirements.

LLM Integration & Agent Development

Building sophisticated AI agents and integrating Large Language Models into existing workflows for enhanced automation and intelligence.

Trading & Financial Systems

Developing AI-powered trading infrastructure, backtesting engines, and quantitative analysis systems with comprehensive performance analytics.

Key Experience

AI-Powered Trading Infrastructure

Built various trading systems from scratch, featuring:

Integration with 10+ financial APIs for market data and intelligence
AI-driven analysis for both fundamental and technical trading strategies using Claude Sonnet and ChatGPT models
Custom SQL indicators for advanced market analysis and signal generation
Vector search implementation using AWS S3 Vectors for semantic document analysis
Custom backtesting engine with performance analytics (Sharpe ratio, Sortino ratio, drawdown calculations)
Real-time data processing using Python and DuckDB for low-latency market analysis
Batch data pipeline orchestration with Dagster for historical analysis and reporting

Trading Report Card Application

Built and maintain Trade Report Card - a sophisticated trading performance tracker implementing Lance Breitstein's methodology with innovative zero-infrastructure architecture:

Innovative URL-based sharing system - entire reports encoded in shareable links without any backend database
Privacy-first architecture with zero server-side data storage - all user data remains in their Google Drive
Google OAuth 2.0 integration with restricted scopes for secure authentication and private data storage
Professional PDF report generation with jsPDF including embedded charts and Google Drive distribution
Automated AWS deployment pipeline with infrastructure-as-code approach using custom bash scripts
Two-tier storage architecture: LocalStorage for immediate access with Google Drive cloud sync
Production code protection using JavaScript obfuscation while maintaining AI agent compatibility
Time-block analysis system tracking pre-market, opening, mid-day, and after-hours sessions
Custom hooks architecture for data management, authentication, and export functionality
Conflict resolution system for data synchronization between local and cloud storage

AI Agent Development

Developed sophisticated AI systems including:

Hierarchical Memory Middleware - Experimental 4-level compression system enabling infinitely long AI conversations, implementing automatic memory management via MCP tools for browsing/expanding compressed conversation history stored in DuckDB
MCP Code Editor specifically designed for AI agent interactions (10+ stars, 5+ forks on GitHub)
Custom agentic workflows using PydanticAI for structured AI interactions
Semantic parsing systems for long documents using LlamaIndex

Data Engineering at TrueRoll (2021-2025)

Built robust data systems for a GovTech startup:

Developed proactive monitoring systems for fraud detection
Maintained comprehensive data pipelines with monitoring and analysis capabilities
Created APIs for data products with focus on integrity and real-time processing

AI & Data Engineering Solutions