Project description
About the Project: Over 5 billion people use traditional, non-AI-native apps daily for email, notes, tasks, and calendars. Our client's mission is to build proactive applications for everyday users who aren't accustomed to complex prompting. Their platform aims to bring intelligence to conversations, errands, organization, and workflows effortlessly. By focusing on persistent context, real-world task execution, and high reliability for long-running workflows, their app minimizes AI hallucinations—with the ultimate objective of organizing users' lives so they can focus on what truly matters.
Technical requirements
- Python
- PyTorch / JAX
- Production ML systems running on GPUs
Responsibilities
As an ML Engineer, you will build core ML components. You will work on real production systems from day one, learning how large-scale ML behaves outside of research settings. This role is for engineers who want to develop strong systems judgment by shipping, debugging, and iterating on real-world ML.
- Build and improve ML components across data, training, evaluation, and inference.
- Fine-tune and adapt models as part of larger production systems.
- Implement evaluation and testing to understand model behavior.
- Help build and maintain data pipelines for real-world and synthetic data.
- Debug model issues, performance problems, and production incidents.
- Ship improvements iteratively and learn from real user feedback.
- Work closely with senior ML engineers, research and product teams.
- Work under real production constraints: latency, cost, reliability, and safety.
- Ensure data pipelines, training loops, and inference systems are robust, reproducible, and maintainable.
Must have
- Strong foundations in machine learning and modern neural architectures.
- Some hands-on experience training, fine-tuning, or deploying ML models.
- Comfortable writing production-quality code and learning new tools quickly.
- Curious, coachable, and eager to learn from real systems in production.
- Able to work through ambiguity with guidance and grow ownership over time.
- Bias toward shipping, iteration, and continuous improvement based on real-world signals.
Recruitment process
Applications are evaluated by our technical team members. Interviews will be conducted via virtual meetings and/or onsite. We value transparency and efficiency, so expect a prompt decision. Please note that due to the volume of applications, we will only contact selected candidates.
Got questions?
To learn more details about this job contact Joanna at [email protected]