Student Researcher Program (Formal Reasoning, LLM Post-Training)

Reasonable is the applied AI research company building formal verification for post-human software development.

Thanks to super-human AI capabilities, code generated by AI can be provably correct, resilient to failures and cyberattacks. At Reasonable, we are doing the research, training the models, and developing the products required to make this a practical reality. Achieving this creates a new paradigm for high accountability software development and unlocks the full potential of AI for professional engineers.

We’re a compact, talent-dense technical team, with deep domain expertise in machine learning, formal verification and software system engineering.

Our interns, like our full time members of technical staff, are outstanding. 10+ international olympiad medals among them, all excellent in their career stages, all thrown in at the deep end and succeeding. Far from working on isolated projects, our interns last year have contributed in a major way to our core research and development.

Join us to develop the next frontier of formal reasoning and software engineering.

Proof follows function.

The Role

As a Student Researcher, you will be embedded within an industry-leading team, working on projects that contribute directly to our work on formally-grounded software development, and/or frontier model evaluations and LLM agent post-training. Our Student researcher programs vary in duration from 12-24 weeks, and we are looking for two types:

- Ambitious Young Students in early years of Undergraduate: typically with a background in competitive programming, maths, or substantial side-projects

- PhD students in Formal Verification and/or Machine Learning

You Will

• Take on a focused research project that connects to Reasonable's core work in machine learning and formal verification. We will shape the project with you on arrival
• Run real experiments that produce results worth publishing or open-sourcing: training runs, verification benchmarks, theoretical work, product development contributions
• Work alongside Ferenc and the wider team of technical staff members in mutual exchange: you bring your area of expertise, eagerness to learn, we bring practical production and research experience
• Be in the room for technical conversations that go beyond your project
• Work on-site, alongside the founding technical team in Budapest or London

• Research Students: Currently enrolled in a PhD or Research Masters programme in Computer Science, Applied Mathematics, or a related technical field, and returning to your program after the internship
• Junior Students: Currently enrolled in early years of undergraduate, or in exceptional cases High School. Demonstrated track record of excellence in maths, programming or AI, such as olympiad results.
• Expertise in a relevant area of computer science, for example: Formal Language Processing, Machine Learning (especially LLMs, RL, other post-training methods), Algorithmic Foundations of Optimization, Mathematical Models of Program Semantics, Programming Language Theory, Design of Distributed Systems, Cybersecurity, or a related field
• Extra-curricular interest in the intersection of machine learning, formal methods, and software engineering, even if your current work doesn’t touch it

Bonus Points If You

• Have published at venues like NeurIPS, ICML, ICLR, POPL, PLDI, CAV, or similar
• Have open-source contributions in relevant projects: LEAN, Verus, Rocq, training frameworks, or similar
• Are already working on the intersection of machine learning and formal methods

This is a serious research program. If you fit the profile, apply. If you know someone who would be ideal, we always reward great introductions.

• Compensation: Generous compensation
• Location: Budapest, Hungary, or London, UK - we’re an on-site team
• World-class team and environment: an opportunity to build alongside deeply experienced founders, in a well-funded company that’s backed by industry leading VCs and angels including Oriol Vinyals, Zoubin Ghahramani, Jonathan Frankle and Guy Podjarny, amongst many others