Title: Build an AI-Powered Anomaly Detection Agent (On-Premise) Description: Looking for an experienced AI/ML engineer (with backend skills) to build a deployable (via Docker) anomaly detection agent. The solution should: The agent should: Ingest data from multiple sources (system/application logs, relational DB tables, basic server metrics like CPU/memory). Detect anomalies using ML/statistical models (e.g., outliers, unusual trends, unexpected errors). Provide alerts in a simple web-based dashboard with basic visualization. Notification hooks: Email + Slack (templated messages in natural language, e.g., “High CPU usage on server-01 for 15 min, trending towards 95% utilization”). Generate natural-language style alert messages (templated or ML-based). Tech preferences: Python for ML, Golang/Python for agent, Scikit-learn/PyOD/Prophet for anomaly detection, lightweight DB for storage, React/minimal UI. Acceptance Criteria: Container builds and runs without errors. Sample datasets trigger correct detection of outliers, DB anomalies, and error bursts. Alerts are delivered to both email and Slack with correct context. False positives, if they occur, will be corrected by admin feedback, and the system will adapt its learning to avoid repeating the same mistakes. Goal: Deliver a feature-rich MVP, with a focus on reliability and adaptability.