Tonic.ai is looking for a hands-on Machine Learning Engineer to help build production-grade NLP systems that power our data privacy and information extraction products. You'll join a small, experienced team working at the intersection of LLMs, data privacy, and applied AI — developing and fine-tuning models that detect and redact sensitive information across diverse datasets.
What You’ll Do
- Build and ship models. Fine-tune and evaluate transformer-based models (e.g., RoBERTa, Gemma, LLaMA) to support PII redaction, entity extraction, and synthetic data generation.
- Own the ML lifecycle. From dataset curation and experiment tracking to model deployment and monitoring — you’ll own the full path from prototype to production.
- Collaborate cross-functionally. Partner with Product and Design to shape how ML models drive user-facing features, and work with the broader engineering team to integrate them into scalable systems.
- Experiment responsibly. Document your experiments, evaluate results rigorously, and help push the frontier of safe and explainable AI for data privacy.
What You’ll Bring
- 3+ years of professional experience in applied ML or data science with a focus on NLP
- Proficiency in Python and deep learning frameworks such as PyTorch and Hugging Face Transformers
- Hands-on experience with experiment tracking (e.g., Weights & Biases), distributed training (e.g., Accelerate), and model serving (e.g., vLLM)
- Comfort working independently and iterating quickly — you enjoy the mix of research, engineering, and product thinking
- Strong communication and collaboration skills
Bonus Points For:
- Experience with supervised and reinforcement learning fine-tuning (e.g. TRL)
- Familiarity with data privacy, PII redaction, or healthcare data
- A public portfolio, blog, or open-source contributions that demonstrate your technical depth and curiosity
Why You’ll Love It Here
- High autonomy and meaningful ownership — your models will ship to production, not sit in a notebook
- Small, collaborative team with deep expertise in NLP and privacy
- Opportunity to work with real-world, high-impact data in domains like healthcare and financial services
About Tonic.ai
Tonic.ai empowers developers while protecting customer privacy by enabling companies to create safe, synthetic versions of their data for use in software development, model training, and AI implementation. Founded in 2018, with offices in San Francisco, Atlanta, New York, and London, the company is pioneering enterprise tools for data transformation, de-identification, synthesis, and subsetting, in pursuit of its mission to make data usable. Thousands of developers use data generated with Tonic.ai on a daily basis to build their products faster in industries as wide ranging as healthcare, financial services, logistics, edtech, and e-commerce. Working with customers like eBay, Cigna, American Express, and Volvo, Tonic.ai innovates to advance its goal of advocating for the privacy of individuals while enabling companies to do their best work. For more information, visit https://www.tonic.ai or follow /tonicfakedata on LinkedIn.
