Application Guide

How to Apply for PhD Studentship: Causal Reinforcement Learning

at Phaidra

🏢 About Phaidra

Phaidra is a deep-tech startup applying AI to control industrial systems like data center cooling and manufacturing, aiming to reduce energy waste and carbon emissions. Founded by AI researchers, the company bridges cutting-edge reinforcement learning research with real-world impact, offering a unique environment for PhD research that directly addresses climate change.

About This Role

This PhD studentship focuses on foundational research at the intersection of causal inference and reinforcement learning, developing algorithms that leverage causal structure for robust generalization. You'll formalize policy learning from biased datasets, benchmark methods on simulated environments, and contribute to AI systems that can safely operate in complex industrial settings.

💡 A Day in the Life

Your day might start with a literature review and coding experiments on a simulated industrial environment (e.g., a data center cooling model). You’d then discuss theoretical insights with your supervisor and collaborate with Phaidra’s engineering team to align your research with practical constraints. Afternoons could involve writing up results, analyzing experiment logs, or presenting progress in a group meeting.

🚀 Application Tools

Generate Cover Letter

AI-powered, tailored for Phaidra

Interview Prep

Practice questions for this role

🎯 Who Phaidra Is Looking For

  • Strong academic background (first-class or 2:1) in CS, Math, Engineering, or Statistics with deep knowledge in at least one of: RL, ML, probabilistic modeling, or control theory.
  • Proficient in Python and PyTorch, with experience building and evaluating RL algorithms (e.g., DQN, PPO, SAC) in simulated environments.
  • Familiar with causal inference concepts (e.g., DAGs, do-calculus, instrumental variables) and their application to machine learning problems.
  • Excellent scientific writer with ability to communicate complex ideas clearly; prior publication in top venues (NeurIPS, ICML, ICLR) is a plus.

📝 Tips for Applying to Phaidra

1

Tailor your research proposal to explicitly connect causal inference and RL, citing specific challenges like distribution shift or off-policy evaluation.

2

Highlight any experience with offline RL or causal discovery in your CV, even if from projects or coursework.

3

Mention familiarity with Phaidra’s industrial applications (e.g., data centers, HVAC) and how your research could improve real-world control.

4

Include a link to your GitHub or personal website with code samples from relevant RL or causal inference projects.

5

In your personal statement, discuss your motivation for doing a PhD at a startup rather than a university—emphasize applied impact.

✉️ What to Emphasize in Your Cover Letter

['Explain why you want to do a PhD in a startup environment and how Phaidra’s mission aligns with your research interests.', 'Describe a specific idea for integrating causal reasoning into RL (e.g., using causal graphs for credit assignment) and its potential impact.', 'Emphasize your technical skills in Python/PyTorch and experience with RL environments (e.g., Gym, MuJoCo, custom simulators).', 'Mention any collaborative or interdisciplinary work you’ve done, as the role involves bridging theory and application.']

Generate Cover Letter →

🔍 Research Before Applying

To stand out, make sure you've researched:

  • → Read Phaidra’s blog posts and research papers on their approach to AI-driven control (e.g., on their website or arXiv).
  • → Study the latest work on causal RL (e.g., papers from ICML 2024 workshops, or authors like Elias Bareinboim).
  • → Explore Phaidra’s industrial use cases: data center cooling, HVAC, and manufacturing—understand the domain constraints.
  • → Review the company’s founding team background (e.g., from DeepMind, Google Brain) to align with their research culture.

💬 Prepare for These Interview Topics

Based on this role, you may be asked about:

1 How would you formalize policy learning from biased observational data using causal inference?
2 Describe a causal structure that could improve out-of-distribution generalization in an RL setting.
3 Walk through a recent RL paper and critique its assumptions about the data-generating process.
4 How would you benchmark a causally-aware RL algorithm against standard baselines?
5 What are the key challenges in applying RL to industrial control systems (e.g., safety, sample efficiency)?
Practice Interview Questions →

⚠️ Common Mistakes to Avoid

  • Submitting a generic research proposal not focused on causal inference and RL—be specific.
  • Ignoring the applied context; don’t treat this as a pure academic PhD—emphasize real-world impact.
  • Overlooking the requirement for strong Python skills; ensure your CV demonstrates coding ability through projects or open-source contributions.

📅 Application Timeline

This position is open until filled. However, we recommend applying as soon as possible as roles at mission-driven organizations tend to fill quickly.

Typical hiring timeline:

1

Application Review

1-2 weeks

2

Initial Screening

Phone call or written assessment

3

Interviews

1-2 rounds, usually virtual

✓

Offer

Congratulations!

Ready to Apply?

Good luck with your application to Phaidra!

← Back to Job Listing Apply Now →