AI Intern – Scalable Dialogue Management for Real‑Time Conversational Systems

AstraZeneca Global Hub is an innovation-driven global biopharmaceutical company focused on the discovery, development and commercialization of prescription medicines, primarily for the treatment of cardiovascular, renal and metabolic, respiratory and immunological, oncological and rare diseases.

In line with our firm commitment to training and developing exceptional talent, this person will join our Enterprise AI Platforms & Services team to learn and grow through active collaboration with our team.

Conditions:

Duration: 6-9 months

Hours per day: Full time

Location: Barcelona, Spain

Focus area:

This internship focuses on the research and development of next-generation duplex and asynchronous conversational AI systems. The goal is to explore and prototype a scalable dialogue management framework capable of supporting multi-threaded, real-time interactions. This includes orchestrating background tasks (e.g., external API calls, computations, knowledge retrieval) while maintaining a smooth and natural flow of conversation. The intern will contribute to designing intelligent dialogue agents that handle overlapping user intents, deferred answers, and dynamic context-switching — pushing the limits of industrial-grade conversational systems.

Training through practical experience:

The internship is structured as a deep-learning experience in advanced dialogue technologies and production-scale AI orchestration.

The intern will:

Support the design of full-duplex conversational loops that enable simultaneous speech input and output.
Assist in implementing an asynchronous task scheduler that manages and resolves long-running background actions (e.g., vector search, computation, external queries).
Collaborate in building a dialogue thread manager that maintains multiple open user intents and gracefully resumes them when responses are ready.
Contribute to real-time testing and evaluation environments simulating multi-turn user sessions with dynamic task handling.
Work under close supervision to gain experience in integrating ASR, NLU, dialogue management, NLG, and TTS modules into a coherent pipeline.

The tasks will be adapted to the intern’s level and focused on learning, exploration of real-world scenarios, and contribution to prototypes and evaluation. This experience is ideal for students looking to bridge research and engineering in conversational AI.

Requirements

Background in Natural Language Processing, Speech Technologies, or Machine Learning
Elegibility for an internship agreement
Proficiency in Python and familiarity with modern AI development practices
Understanding of conversational system pipelines, including components like speech recognition, language understanding, and dialogue management
Interest in multi-tasking dialogue systems, scalable AI orchestration, and human-like conversational flow
Motivation to engage in both research-driven experimentation and practical system design
Bonus: experience with large language models (e.g., GPT, Claude, LLaMA), streaming architectures, or open-source dialogue frameworks.

#EarlyTalent

AstraZeneca is a company that promotes equal opportunities. AstraZeneca will consider all eligible persons, without discrimination based on disability, sex or sexual orientation, pregnancy or maternity/paternity period, race or origin (national or ethnic), age, religion or belief, identity of gender, marital or de facto partnership status and/or any other characteristic protected by law.