View Uppdrag

Apply using URL

General

Headline

Organization/Company

Location

Assignment Type

Description

This degree project offers a unique opportunity to join our pioneering startup in Stockholm’s vibrant tech scene. You’ll work side-by-side with our CEO, CPTO, and technical team in our premium office in downtown Stockholm, under the direct supervision of the founding data scientist.

As part of our core team, you’ll help redefine how AI training data is valued - keeping pace with the ambitions of leading model trainers and agent builders. The emerging field of data valuation in AI is rapidly evolving. It enables model developers to identify the most effective data for training, while giving data owners insight into how their content influences model performance. For content creators, accurate valuation and pricing are essential to ensure proper compensation when proprietary or copyrighted materials are used in AI training. Redpine is pioneering in this space by leveraging the latest techniques from data valuation in our AI data platform https://www.redpine.ai/. 

Your work will span the full spectrum, from engaging with the most impactful research published at venues such as NeurIPS, to translating cutting-edge algorithms into our proprietary platform through hands-on engineering. This thesis combines academic rigor with real-world execution, demanding both scientific depth and product-oriented thinking.
Expect to explore, test feasibility of and implement novel data attribution techniques and pre-valuation algorithms, such as [1], [2] and [3], and contribute directly to client-ready tools. This is not a traditional academic project but an invitation to push boundaries at the intersection of research and entrepreneurship.
At Redpine, you’ll join a fast-moving, collaborative environment where your ideas can shape executive strategy, guide technical development, and influence our live systems. It’s a living laboratory for those eager to think deeply, challenge frontier research, and build tangible results alongside the leaders defining this new domain.

What we look for:
- Strong scientific mindset
- Proficiency in Python and basic ML, and NLP/LLM concepts
- Experience fine-tuning or training language models is a plus
- Entrepreneurial mindset
- Desire to make tangible impact


About Redpine
Redpine powers AI builders and autonomous agents with access to licensed, high-quality, non-public and multi-modal data - securely and at scale. Founded in 2024 and based in Stockholm, Redpine is backed by investors from OpenAI, Spotify, Perplexity, Xiaomi, Sana, and more.

Resources:
[1] Enhancing Training Data Attribution with Representational Optimization. Weiwei Sun and Haokun Liu and Nikhil Kandpal and Colin Raffel and Yiming Yang https://arxiv.org/pdf/2505.18513v1

[2] What is Your Data Worth to GPT? LLM-Scale Data Valuation with Influence Functions. Sang Keun Choe and Hwijeen Ahn and Juhan Bae and Kewen Zhao and Minsoo Kang and Youngseog Chung and Adithya Pratapa and Willie Neiswanger and Emma Strubell and Teruko Mitamura and Jeff Schneider and Eduard Hovy and Roger Grosse and Eric Xing
https://arxiv.org/abs/2405.13954

[3] Estimating Training Data Influence by Tracing Gradient Descent. Garima Pruthi and Frederick Liu and Mukund Sundararajan and Satyen Kale
https://arxiv.org/abs/2002.08484

There are no records to display.

You don't have permissions to view these records.

Error completing request.

Deadline for application

Publish date

Credits

Application Channel

Application Email

Application URL

Application Documents

CV, transcripts