AI Scientist

Neuralk; Paris; Hybrid; EUR 100,000 - 140,000; Full time

Benefits offered by this job

Equity (BSPCE)

Paid leave

Flexible remote work

Job summary

Neuralk is a deep-tech startup building the next generation of Foundation Models for Data Science. Our Data Science agent helps data scientists from problem framing to production-ready models, focusing on structured data.

Reporting to NicolaCancedda, VP Research, the role is based in Paris or London with in-person collaboration and occasional remote work. You will join a fast-moving, research-driven environment with ownership and the opportunity to shape a core product.

Qualifications

• PhD in Computer Science, ML or related field required.
• 3+ years of ML experience including pre-training, fine-tuning and evaluation of DL models (Transformers).
• Publication record in top ML conferences or journals.
• Excellent English communication.
• Experience with large-scale ML experiments and cloud/private clusters.

Responsibilities

• Develop foundation models for structured data (tables/time-series).
• Collaborate with ML engineers to improve data science agent performance.
• Design and evaluate metrics reflecting client use-cases.
• Implement active learning strategies and data optimization for training datasets.
• Present ML research concepts to the team and stakeholders.
• Collaborate with clients to deliver effective representations.

Skills

Transformers

Python

C/C++

Research

Communication skills

Autonomy

Interdisciplinary teamwork

Education

PhD in Computer Science or Machine Learning

Tools

PyTorch

Deepspeed

SLURM

Job description

Neuralk-AI is looking for an AI Scientist with experience in AI model design and training.

You must have a PhD to apply.

You must have validated one of the following options below:

• you have pre-trained large transformer-based models -> pre-training team
• you have fine-tuned large transformer-based models -> fine-tuning team
• you have worked on model training and data distribution modelling -> synthetic data modelling team

You will report to the NicolaCancedda (VPResearch) and will be located in our Paris or London offices.

Neuralk is a deep-tech company building the next generation of Foundation Models for Data Science. Our mission is to build the predictive layer for businesses, transforming data science from a series of one-off initiatives, stitched together across silos, overly bespoke, and dependent on a handful of specialists, into a durable capability: a scalable predictive infrastructure that continuously learns from an organization’s data and powers decisions across the enterprise.

Our product is a Data Science agent, powered by our Foundation Models, that assists data scientists throughout their workflow, from problem framing to robust, production-ready models. We focus on the hardest and most common data problems in companies: structured datasets describing customers, operations, risks or financial activity.

As an early-stage, well-funded AI startup, Neuralk builds on state-of-the-art research to solve concrete business challenges. We value clarity over complexity, strong fundamentals over hype, and fast iteration grounded in rigorous engineering. Our ambition is to redefine how predictive AI is built and used in organizations, at scale.

Joining Neuralk means working hard in a fast-moving, research-driven environment, with a high level of ownership and the opportunity to shape a core product at the intersection of machine learning, engineering and real-world impact.

Mission Highlights:

As a Machine Learning Researcher, your role will be to contribute of the development of our foundation models for structured data (table and time-series). You will collaborate closely with our engineering team (~8 people) to enhance the performance, scalability and impact of our data science agent.

Role & Responsibilities:

By contributing to the core of our AI research effort, you will be responsible for:

• Algorithms: Contribute to the development of foundation models for structured data.
• Evaluation: Continuously evaluate and optimize the performance of our models by building adapted metrics reflecting the use-cases of our clients, building upon the insights from our industrial and academic partners.
• Active learning and training data optimisation: Participate in the active learning strategy and implementation process to improve sample selection and future model performance. As well as designing and consolidating training and evaluation datasets to optimise representational as well as transfer learning abilities of our Tabular Foundation models.
• Research: Stay current with the latest ML advancements in the field and suggest optimisations that may improve the foundation models’ performance and capabilities.
• Pitching & communication: present both ML research concepts to the scientific community and experimental design needs to the ML team.
• Collaboration: Work closely with ML engineers, data scientists, and clients to deliver promising representation algorithms for downstream applications.
• Ad-hoc analyses: Running analyses to understand the learning mechanisms of the foundation model.

Profile:

• PhD in Computer Science, Machine Learning or a closely related field, with a focus on deep learning.
• 3+ years of experience in machine learning which involved pre-training, fine-tuning and evaluating DL algorithms (Transformers) in the cloud or in a private cluster.
• You have a publication record in top-tier ML conferences or journals
• Excellent communication skills in English.
• Proven ability to work with interdisciplinary teams.
• Thrives in a fast-paced, evolving startup environment.
• Self-starter and autonomous.
• Strong analytical skills and problem solving ability.
• Appetite to explore, implement new ideas and innovate.
• You have demonstrated experience in designing and running large-scale ML experiments (SLURM, Pytorch, Deepspeed).
• Demonstrated machine learning experience in one of the following: open-source activity, data science competitions.
• Track record of translating research into business impact
• Experience in developing and debugging in C/C++, Python

Expertise:

• Machine Learning: Deep understanding of ML theories and practices, especially related to reproducibility and scalability.
• Foundation Models: Experience in designing, training and evaluating large-scale foundation models.
• Programming: Proficient in Python and AI frameworks and tools (e.g., Sklearn, PyTorch), with experience in software development best practices and version control systems such as Git.
• Data management: Familiarity with data structures and database systems (Parquet, SQL and NoSQL), to manage and process large datasets efficiently.
• AI platforms: Experience with deploying and managing machine learning models, including familiarity with Pytorch, SLURM and Deepspeed (or similar).

We are a fast-pace startup, yet, we favor a good work-life balance and interesting compensations. We offer:

• Equity (BSPCE), to reflect the value you bring to Neuralk and to foster a shared journey
• French level paid leave and time-off work
• Dynamic work setting. Although our preference is for in-person collaboration, we will be flexible with occasional remote work arrangements.