Machine Learning Engineer for Translation/LLM (Remote with Los Angeles Visit)

Job Details

This is a remote position.

Job Title

Machine Learning Engineer (C++/Python) for Translation/LLM (Remote with Los Angeles Visit)

Job Description

Codesta, an IT Staff Augmentation company, is currently seeking a Machine Learning Engineer to join our innovative client project based in Los Angeles. The ideal candidate will play a crucial role in enhancing and optimizing machine translation models for a range of languages, focusing on improving translation quality and operational efficiency.

Key Responsibilities

• Model Optimization: Fine-tune custom translation models using C++ and Python to ensure high translation quality across various languages, with a particular focus on addressing the unique challenges presented by non-Latin-based languages.
• Data Management: Leverage Marian and work with a diverse corpus of data (Opus-MT), focusing on the acquisition and enhancement of high-quality data for languages with limited source material. Utilize Marian's capabilities to optimize the processing and integration of this data for improved translation outcomes.
• API Integration: Integrate individual language models with the client’s central API to ensure seamless data flow and translation output.
• Quality Assurance: Implement strategies to bring the quality of translations up to a high standard, addressing specific issues encountered with certain languages.
• Operational Efficiency: Advise on setting up a production line for at-scale translations, optimizing the process for efficiency and effectiveness.
• Collaboration and Communication: Work closely with a small, technically adept team within a large organization, providing expertise in machine translation and natural language processing to support ongoing experimentation and development.

Requirements

• Technical Expertise: Familiarity with C++ for core model understanding and strong Python understanding for working with the translation framework. Experience with machine translation and natural language processing is essential.
• Experience: Proven track record of improving the quality of machine translation models, with specific experience in dealing with languages that have quality issues.
• Problem-Solving: Creative and analytical approach to problem-solving, with the ability to work on complex challenges and deliver high-quality solutions.
• Communication: Excellent communication skills, with the capacity to advise on technical matters and operational efficiency to a team with varying levels of technical familiarity.

Additional Information

• Interview Process: Codesta will conduct initial testing for technical and soft skills. Pre-screened candidates will be interviewed by the client via Zoom.
• Location: Remote, but a face-to-face meeting lasting for a few days is required in Los Angeles. All-inclusive travel covered. The rest of the project can be done remotely.
• Travel: Since face-to-face time in Los Angeles is required, the candidate must be able to travel to the United States without restrictions or delays.
• Duration: The project is short-term, estimated to last up to a month, with potential for further consultation down the line.
• Compensation: Negotiable.