The position requires a minimum of 6 years of experience.
The job is located in the USA and is a full-time position.
The role involves owning the innovation pipeline, transforming research into real-world impact, and building unmatched systems in the market.
Responsibilities include inventing and implementing cutting-edge advances in LLMs, NLP, OCR, and computer vision to apply them to complex, multilingual, high-stakes financial documents.
The candidate will own the model lifecycle, leading end-to-end development including data collection and curation, architecture design, training, evaluation, and deployment.
The role requires designing and refining internal evaluation frameworks that set new standards for model quality in financial data extraction.
The candidate will rapidly build and ship experimental features while maintaining scientific rigor.
Collaboration with product, infrastructure, and leadership teams is essential to scale innovations into production systems for global financial use cases.
Requirements:
The candidate must demonstrate the ability to translate new research into working prototypes within days, not months.
Strong Python skills with deep learning frameworks, preferably PyTorch, are required, along with the ability to debug and productionize their own work.
Proven experience in training and optimizing large models for NLP, computer vision, or multimodal tasks is essential.
The candidate should be skilled at handling messy, real-world datasets, not just curated academic benchmarks.
An end-to-end problem solver who thrives on responsibility and is not limited to isolated pipeline tasks is necessary.
Benefits:
The role offers immediate impact, allowing the candidate to see their research deployed to production within weeks.
The candidate will have ownership of projects from inception to deployment with no silos.
The position allows for a faster pace than big-tech labs while maintaining research excellence.
The candidate will gain visibility by working directly with senior leadership and decision-makers.
The role provides an opportunity to help shape the technical foundation of a next-generation AI platform.