IA-01 Design of an Advanced Cognitive Assistant for Interactive Multidimensional Analysis and Intelligent Synthesis of Contextualized Insights within a Decision Dashboard PFE

Project Overview

• Design and implement an advanced cognitive assistant embedded within a decision dashboard to support interactive multidimensional analysis.
• Provide intelligent synthesis of contextualized insights, enabling users to query, explore and obtain concise, actionable summaries from complex datasets.

Responsibilities and Tasks

• Research and prototype NLP pipelines for contextual understanding, entity extraction and intent classification using libraries such as spaCy and NLTK.
• Integrate open-source generative AI components to produce human-readable syntheses and recommendations tailored to dashboard contexts.
• Collaborate on backend integration in Java and ensure persistence of insights and metadata in PostgreSQL.
• Develop frontend components in Angular to support interactive query interfaces, visualization hooks and user feedback loops.

Technologies & Technical Expectations

• Use Python for AI model development, training and orchestration; employ spaCy/NLTK for linguistic processing and open-source generative models for synthesis.
• Ensure backend interoperability with Java services and design efficient storage schemas in PostgreSQL for insights, logs and contextual data.
• Build Angular UI components that allow multidimensional filtering, natural-language querying and display of generated summaries; consider performance and UX for interactive analysis.

Deliverables & Evaluation Criteria

• A working prototype of the cognitive assistant integrated into the decision dashboard demonstrating: contextual query handling, multidimensional analysis flows and generated insight summaries.
• Documentation covering model choices, data preprocessing, API integration details, frontend usage instructions and evaluation metrics (accuracy, relevance, latency).
• Unit/integration tests for core modules and a small dataset/case studies used to validate and demonstrate system capabilities.

Candidate Profile & Skills

• Strong experience in Python for NLP and model development; familiarity with spaCy, NLTK and open-source generative AI is expected.
• Knowledge of Java backend integration practices, RESTful APIs and PostgreSQL database design.
• Frontend competence with Angular to implement interactive UI features and connect to backend services.
• Good analytical skills to translate decision-making needs into model requirements and UX interactions.

How to Apply

• To apply for this specific project, send your CV and a short motivation letter to stages@hydatis.fr.
• For any questions about the IA-01 project, use the same contact email to request further details or to propose a short meeting.