Member of Technical Staff, Model Training

Parallel; San Francisco (CA); On-site; USD 120,000 - 150,000; Full time

Benefits offered by this job

Competitive salary

Generous equity

Visa sponsorships

401K plans

Daily lunch & office snacks

Dinner at the office

Unlimited vacation

Caltrain pass reimbursement

Job summary

Parallel is seeking a professional who will own the training pipeline behind models that support both the search stack and agents. Responsibilities include building pathways from product usage to high-quality training data, rigorously fine-tuning models, and shipping them for high-traffic usage.

The ideal candidate should have a strong grasp of modern training models, especially in the realms of transformer architectures and data curation, ensuring applied research benefits millions of users.

Qualifications

• Deep intuition on modern models and training including transformer fine-tuning.
• Experience with data curation and ensuring label quality.
• Strong understanding of how ranking, retrieval, and agent behavior inform one another.

Responsibilities

• Own the training pipeline behind models for search and agents.
• Build the path from product usage to high-quality training data.
• Fine-tune and evaluate models rigorously, ensuring they are safe for traffic.

Skills

Deep intuition on modern models and training

Transformer fine-tuning

Data curation

Label quality

Understanding of ranking and retrieval

Job description

Job

You will own the training pipeline behind the models that power both Parallel’s search stack and Parallel’s agents. On the search side, that means the rankers, classifiers, and query models that surface the right information. On the agent side, that means the models that help our agents plan, reason, and execute high‑value tasks over web data. You will build the path from real product usage to high‑quality training data, fine‑tune and evaluate these models rigorously, and ship them safely to traffic used by millions.

Qualifications

Have deep intuition on modern models and training, including transformer fine‑tuning, data curation, and the craft of label quality. Think rigorously about how ranking, retrieval, and agent behavior inform one another, and how to train models that serve all three. You care about your research being applied to product and systems that millions use.

Compensation & benefits

• Competitive salary
• Generous equity
• Visa sponsorships
• 401K plans
• Daily lunch & office snacks
• Dinner at the office
• Unlimited vacation
• Caltrain pass reimbursement