Responsibilities
- Model Development and maintenance: Develop, iterate, and maintain models to assist humans in providing information about data points for training and evaluating large scale foundation models. Build models across modalities (Audio, Text, Video) to assist labelers in their tasks and automate labeling where possible.
- Clarification and Simplification of Model Requirements: Diligently comprehend machine learning development lifecycle, drive clear communication on where most impact can be made, and decompose technical tasks into more manageable units, ensuring a smooth execution path.
- Explore agentic multimodal approaches to improve data quality for model training, evaluation, and feedback. Augment human labeling with synthetic data and automated labeling.
- Collaborative Experimentation and Documentation: Work in tandem with various teams to formulate meticulous experiment plans, documenting results with utmost precision, and patiently addressing inquiries from both technical and non-technical stakeholders.
Basic qualifications
- Bachelor's/Master's degree in Computer Science, Computer Engineering, or other relevant majors.
- 4+ years of Machine Learning experience: Demonstrated previous experience owning the end-of-end machine learning model development process. Understand how data plays a crucial role in terms of model quality, model iteration and model evaluation.
- 4+ years of experience working with Python/Pytorch: Possesses advanced skills in Python programming, showcasing the ability to develop and implement robust models in production environments solutions.
- Statistical Knowledge: Demonstrates strong statistical background and can leverage statistical knowledge in solving business analytical requirements.
- Strategic Interpersonal Skills: Displays formidable interpersonal skills, embodying an ability to seize new opportunities, be able to set product directions strategically, ignite innovative ideas, and marshal support to bring products from conceptualization to realization.
- Familiar with the technical principles of modern LLM/MLLM development and application for content generation and content understanding purposes.
Preferred qualifications
- Creative thinking and a passion for innovation.
- Experience with the AIGC modeling process: Have previously trained or applied MLLM models (T2I and T2V) for key business use cases.
- Understand the data curation process and model evaluation criteria when developing AIGC models.
- Interacting and occasionally having unsupervised contact with internal/external clients and/or colleagues;
- Appropriately handling and managing confidential information including proprietary and trade secret information and access to information technology systems; and
- Exercising sound judgment.
Tags & focus areas
Used for matching and alerts on DevFound Ai Machine Learning