Research Engineer | Reasoning for Agents
About the Role
Rippletide is building the reasoning infrastructure for autonomous agents. We are opening a Ms/PhD research engineer position for someone who wants to work on hard problems at the intersection of language models, reasoning, and decision-making, with direct impact on real-world systems.
You will be embedded in Rippletide's R&D team in Paris, working alongside researchers and engineers on projects that push the boundaries of agent reasoning. Your work will be grounded in real data and real business problems, and you will have the opportunity to publish and co-author papers or demos based on your contributions.
Rippletide is based in Paris and San Francisco.
Project Examples
1. Multi-Agent Reasoning Systems
Work on improving how autonomous agents cooperate and align their objectives to achieve complex, shared goals. The project focuses on designing alignment mechanisms, coordination strategies, and communication protocols that enable agents to reason collectively and adapt in dynamic environments.
Research focus: cooperative reasoning, alignment through shared representations, decentralized decision-making, and emergent coordination behaviors.
Ideal profile: Ms/PhD student with expertise in multi-agent systems, reinforcement learning, distributed AI, or computational game theory.
2. Model Predictive Control for Decision Optimization
Work on how predictive models can reason over uncertainty, dynamically adapt control strategies, and maintain stability and optimality when faced with conflicting objectives or incomplete information. The application involves large and evolving sets of real-world business constraints.
Research focus: large-scale constrained optimization, adaptive MPC, hybrid control architectures integrating learning-based models, and efficient real-time solvers for high-dimensional systems.
Ideal profile: Ms/PhD student with a strong background in control theory, optimization, or applied mathematics.
3. Neuro-Symbolic Reasoning
Work on improving the combination of adaptability from neural networks and the structured reasoning of symbolic systems in neuro-symbolic architectures. The project focuses on building frameworks to evaluate and refine these hybrid models in complex, evolving environments, enabling explainable and generalizable reasoning across tasks.
Research focus: integration of neural and symbolic representations, reasoning under uncertainty, simulation-based evaluation of cognitive architectures, and transfer from synthetic to real-world contexts.
Ideal profile: Ms/PhD student with experience in neuro-symbolic AI, computational modeling, or probabilistic reasoning, and a strong interest in the intersection of learning, logic, and simulation.
What You'll Do
- Collaborate with Rippletide's R&D team on reasoning architectures
- Work on real-world problems and data at the intersection of language models, reasoning, and decision-making
- Publish and co-author papers or demos based on joint work
Location
Paris.