Project overview

• Design and implement AI agents that autonomously tune database parameters and configurations to improve performance.
• Research and prototype novel approaches for AI-driven database optimization across multiple database engines (PostgreSQL, MySQL, MongoDB, etc.).

Primary responsibilities

• Build reinforcement learning models that optimize database performance over time and create feedback loops allowing agents to learn from performance outcomes.
• Implement multi-agent systems coordinating indexing, query optimization, and resource allocation; analyze agent recommendations and validate them against baseline performance.

Technical tasks & methods

• Develop agents that analyze query execution plans and recommend improvements, and profile database internals to identify performance bottlenecks through monitoring and analysis.
• Create predictive models for workload forecasting and capacity planning; benchmark and validate agent recommendations using realistic workloads and performance metrics.

Expected deliverables

• Prototypes of AI agents (single- and multi-agent) capable of tuning parameters and recommending optimization actions, with experimental results comparing to baselines.
• Documentation and reproducible benchmarks demonstrating improvements, plus code and models for reinforcement-learning-based tuning and workload forecasting.

How to apply

• To apply, send your application referencing this specific project to jobs@guepard.run or use the online form at https://lnkd.in/d8wuxwT3.
• Include CV, brief description of relevant experience (databases, RL, profiling), and subject line: "Application: 5 AI Engineer AI Agent for DB-Tuning PFE — Internship".