Member of Technical Staff - Post-training (Data Infrastructure)

Description

The company is a security-first enterprise AI firm that builds foundation models and complete products aimed at practical business problems.

It trains and deploys frontier models for businesses that are building their own AI systems.

The team includes researchers, engineers and designers, and the company is headquartered in Toronto with offices in London, New York City, San Francisco, Montreal, Paris, Berlin and Seoul.

In this role you will build and own the pipelines that generate post-training data for large language models, from SFT and RL data to synthetic and human-annotated data.

You will work alongside researchers and annotation teams so they can revise data as quickly as they revise models.

Responsibilities

• Build and run pipelines that produce SFT, preference and RL post-training data.
• Take ownership of dataset quality, versioning and reproducibility.
• Convert research data recipes into robust, reusable production components.
• Expand model inference so it can generate data and grade outputs with a model.
• Define data formats and schemas that keep post-training datasets consistent and efficient to train on.
• Work with researchers and engineers so your pipelines stay in step with the needs of cutting-edge language models.

Requirements

• Solid software engineering ability.
• Experience designing and building large-scale machine learning data pipelines, ideally for LLM post-training.
• A record of running distributed data processing and model inference at scale.
• A good grasp of post-training methods such as SFT, RLHF and DPO, and of how data influences model behaviour.
• Strong communication skills.
• This list is not exhaustive. If none of it fits perfectly but you believe you can contribute, apply anyway, since the company values varied backgrounds and viewpoints.

Conditions

This is a full-time position.

Remote work is supported, with offices in Toronto, London, New York City, San Francisco, Montreal, Paris, Berlin and Seoul and more planned.

Office-based staff get a daily lunch program, snacks and regular community and social events.

Remote staff who are not near an office get a co-working benefit to work alongside others in their city.

AI-enabled tools may help recruiters screen applicants against this role's criteria, but they do not limit which applications recruiters review.

The company welcomes applicants from all backgrounds and is committed to equal opportunity.

Benefits

• Full-time employees receive a weekly lunch stipend of $75 or £75, or the equivalent in their local currency.
• Health and dental coverage, plus a separate mental health budget.
• RRSP matching, a 401K and a pension scheme.
• A 100% parental leave top-up for up to 6 months, available to either parent.
• Annual enrichment benefits covering arts and culture, fitness and wellness, quality time, and a credit toward workspace improvements.
• A stipend for conferences, courses and coaching.
• 6 weeks of paid vacation, which is 30 working days.
• A travel budget to visit other offices for remote staff, plus an annual company offsite.
• A $500 home office stipend.