AI Quality Analyst, Personalization (Russian)

About OpenTrain

OpenTrain is the #1 platform for building careers in AI training and data labeling. We connect contributors with high-quality, remote projects that let you grow a durable freelance career teaching AI how to behave.

We support flexible, part-time work that fits around your life and help people build a unified portfolio of AI training experience.

About AI training work

AI training (data labeling / annotation / human feedback) is the human side of modern AI: people create, test, and evaluate examples that teach models to respond accurately and helpfully. This work is remote, often entry-level friendly, and lets you directly shape how cutting-edge systems behave.

• Flexible, remote work—fit it around studies, another job, or family.
• Many projects need no prior experience; language fluency and attention to detail are often enough.

The role

As an AI Quality Analyst (Personalization) you will evaluate how well a conversational model uses personal account data to produce grounded, integrated, and helpful responses in Russian. You'll join a global, around-the-clock evaluation team and contribute structured ratings and written rationales that improve personalization quality.

• Type: Contractor, Part-time
• Hours: 20+ hours/week
• Rate: $15 USD per hour
• Language: Russian (reading and writing required)
• Work location: Fully remote, worldwide

What you'll do

Perform structured, side-by-side evaluations of conversational AI responses using multi-turn prompts that rely on your personal account history (past conversations, emails, search and video activity). Carefully compare two model outputs and provide clear, structured rationales.

• Design and run multi-turn prompts (typically 1–5 turns) that require the model to leverage personal context.
• Evaluate responses for Grounding, Integration, and Helpfulness using provided guidelines.
• Stack-rank two model responses and write concise, evidence-based rationales for differences.
• Delete evaluation conversations after each session to maintain strict data hygiene.

Requirements

You must be proficient in reading and writing Russian and able to work independently from a desktop or laptop with a reliable internet connection. This is an entry-level friendly role but expects strong analytical thinking and attention to detail.

• Fluent in Russian (reading and writing).
• BS/BA degree or equivalent in a relevant analytical field (e.g., Linguistics, Computer Science, Journalism, Ethics, Policy, Law).
• Experience in data annotation, AI evaluation, or content moderation.
• Skilled at creative prompt design and comparative analysis of model outputs.
• Excellent written communication for clear rationales and documentation.

Who should apply

Apply if you want flexible remote work that puts you on the front lines of how AI models learn to personalize responses. This role suits careful analysts, language experts, and early-career annotators who enjoy prompt engineering and written critique.

• Ideal for entry-level annotators with prior evaluation or moderation experience.
• Good fit for people who enjoy structuring comparisons and writing clear, evidence-based feedback.

How it works

You will receive task guidelines, rating rubrics, and examples before starting. Work independently on assigned batches, submit ratings and written rationales through the OpenTrain evaluation interface, and follow mandatory data-hygiene steps after every session.

• You will design prompts using your own account data (emails, search history, past conversations, video activity) as specified by the task.
• After each evaluation session you must delete the conversation data per instructions.
• Pay is hourly at $15 USD; you will be contracted and paid according to OpenTrain procedures.