Project overview

• Build an AI Assistant that automatically analyzes error logs, stack traces and exceptions produced by Proxym applications.
• The assistant must explain the probable cause of the bug in French, propose concrete solutions with code examples, and detect recurring error patterns for historical analysis.

Objectives & deliverables

• Implement an analysis pipeline that consumes logs from Elasticsearch/Logstash, enriches them via AI (OpenAI / Claude), and stores results for dashboarding.
• Deliver an IntelliJ plugin for in-IDE explanations and suggestions, plus a React web dashboard for historical error analysis and pattern detection.

Core responsibilities

• Develop backend services (Spring Boot) to process logs, call AI APIs, and expose results via APIs.
• Integrate with Elasticsearch (logs), Logstash, Redis cache, and PostgreSQL for storage and retrieval; implement caching and indexing strategies.

Technologies & tools

• AI / LLM: OpenAI API and/or Claude for natural language analysis and code suggestion.
• Backend: Spring Boot; Database: PostgreSQL; Cache: Redis.
• Observability & logs: Elasticsearch + Logstash; Frontend: React dashboard.
• IDE integration: IntelliJ Plugin SDK to provide inline explanations and code suggestions inside the IDE.

Required profile

• Engineer profile / Bachelor’s degree; position for 1 Trainee.
• Strong Java/Spring Boot skills, familiarity with REST APIs, and interest in LLM integration and developer tooling.

How to apply

• To apply for this specific project (REF: PRX-2026-04), submit your application through the trainees platform: https://trainees-platform.proxym-group.net.
• In your application, highlight past projects with logs/error analysis, experience with LLMs, and any IntelliJ plugin or developer tooling work.