Research Associate - AI

San Francisco, California, United States - Remote

Keywords Studios Plc

Whatever the scale of your project, whatever the timeframe, whatever your location: we deliver the unified solutions you need. Partner with us today.

View all jobs at Keywords Studios Plc

Apply now Apply later

Work Location: USA Remote (Prefer PST Time Zone)

Work Hours: M - F, 9:00 am - 5:30 pm

Pay Rate: $22 - $27/hr DOE and location

The Research Associate is a non-traditional role in which you will work on our data collection and quality team focusing on the improvement of an AI engine.

In addition to an ability to write clearly and concisely, successful Research Associates must be able to tailor their writing style to each assignment’s requirements, must possess solid research skills and be able to quickly paraphrase their findings, and will be called upon to evaluate Large Language Model (LLM) prompts written by others

The ideal candidates will have a solid ability to focus on efficiency and problem-solving, and excellent writing and reading comprehension skills – including experience in creating and composing text within a specified amount of time

This role provides opportunities for career advancement.

Note: There is no coding or software development as part of this role.

What You'll Achieve (Responsibilities): 

  • Work collaboratively in a fast-paced environment
  • Work on various client projects to train generative AI models, by creating prompts and responses based on the instructions provided and on using established best practices for quality prompts
  • Given examples, generate similar prompts and responses
  • Execute different Use Cases collecting Data in support of AI engine
  • Fine-tune AI training prompts for more consistent results
  • Work with a small team of Data Specialists in annotation and labeling projects.·
  • Engage and assist in regular team training
  • Help identify areas for process improvements
  • Assist in documenting processes
  • Quantitative and Qualitative feedback
  • Provide feedback on tools being used and on potential alternatives
  • Use a variety of communication channels such as Slack, Teams, and SharePoint, to learn about new projects, collaborate with your team, and ask questions
  • Learn new software programs on the job
  • Providing supporting documentation when the AI fails

Keywords provides a competitive compensation package, good benefits and a casual, fun, productive and supportive working environment. We empower people to perform to the best of their ability with our “can do” attitude. We appreciate and embrace flexibility and learn at every opportunity to grow ourselves through experience, training and tackling new challenges. This is what makes us Keywordians.

Requirements

  • A candidate with a college degree in Computer Science, Math and Statistics, or Cognitive Science.
  • Preferred prior work experience or college studies in AI
  • Process-oriented, focused on problem-solving, an effective communicator, efficient, and highly organized, with strong attention to detail
  • An ability to learn, document, and work with the team on new technologies and processes
  • Ability to gain new skills and knowledge through hands-on experience
  • Experience in assisting in system troubleshooting & finding resolutions
  • Keen eye for detail
  • Strong Communication Skills (Oral and Written)
  • Demonstrated ability to work independently
  • Strong time management skills
  • Exemplify the quality of doing "get it done attitude," including a high level of accountability, transparency, and teamwork first & foremost

Benefits

At KeyWords we provide all our contingent workforce with:

  • Paid Time Off (including sick days and holidays)
  • 401k (3% matching)
  • Medical, Dental and Vision benefits
Apply now Apply later
Job stats:  8  0  0

Tags: Computer Science Generative AI LLMs Mathematics Research SharePoint Statistics

Perks/benefits: Career development Competitive pay Health care

Regions: Remote/Anywhere North America
Country: United States

More jobs like this