Software engineering subject-matter experts bring technical rigor to AI-training work. On OpenTrain you’ll apply coding knowledge, system design sense, and testing discipline to tasks like annotating code and logs, evaluating model outputs for correctness, and designing datasets and schemas that reflect real-world software behavior. These roles are typically remote and project-based, and they reward people who can read code quickly, write clear guidelines for annotators, and translate software requirements into precise labels and tests.
Join a long-term remote AI training project writing and debugging C# code to build training datasets for models; flexible 20–40 hrs/week at $9/hr. Requires a CS/engineering degree and at least 2 years of professional C# experience; a coding assessment follows application.
Posted Jun 22, 2024
Join a long-term, remote project writing and debugging Go code to create training data for AI models; contract role, 20–40 hrs/week, $9/hr. Candidates must have at least 2 years of professional Go experience and be comfortable working in English.
Posted Jun 22, 2024
Remote contractor role reviewing Swift/iOS code used to train AI models. $14/hr, ongoing 3–6 month project; expect 20+ hours/week (many contributors work 30–40 hrs); must have 5+ years Swift experience and B2+ English.
Posted Jun 18, 2024
Join an ongoing contractor role creating and debugging Python code to train AI models; commit 20–40 hours/week for 3–6 months at $8/hr USD. Ideal for Python developers with 2+ years' experience, strong English, and LLM training experience preferred.
Posted Jun 18, 2024
Join a remote, part-time contract role creating Swift coding prompts and responses to train LLMs — 20+ hrs/week at $12/hr. Must be highly proficient in Swift with 2+ years of iOS experience and strong English; LLM training experience is a plus.
Posted May 17, 2024
Software engineering work in AI training focuses on the technical parts of building and evaluating models that understand, generate, or reason about code and software systems. Typical tasks include labeling code snippets and API traces, classifying bug reports, evaluating whether model-generated code is correct and secure, creating edge-case test cases, and annotating software documentation or logs for downstream training.
Projects also ask engineers to design annotation schemas, write unit-style checks for dataset quality, and produce clear instruction sets that non-technical annotators can follow. Your output helps models learn correct behavior, spot bugs, follow APIs, and produce maintainable code.
Success in these roles draws on practical software-engineering skills rather than academic ML research. Comfortable reading and reasoning about code, debugging, writing tests, and using version control are all highly relevant. Familiarity with common languages and ecosystems (for example, scripting languages, web APIs, or SQL) makes it easier to judge correctness and edge cases.
Equally important are communication and instructional skills: many projects require you to write clear criteria, examples, and counterexamples so other annotators can apply labels consistently. A background in QA, code review, SRE, or technical writing is often transferable.
These roles suit professional software engineers who want flexible, remote work that leverages their technical judgment. They’re also a fit for QA engineers, technical leads, documentation authors, and advanced students who want to apply hands-on coding knowledge without committing to full-time product development.
People who do well enjoy pattern recognition, breaking ambiguous requirements into testable criteria, and mentoring or reviewing others’ work. If you like reproducing tricky bugs, designing edge cases, or explaining why a piece of code is wrong, this facet of AI-training work can be a strong match.
On OpenTrain you build a profile that highlights your technical skills and relevant examples. Many software-engineering projects require a short qualification task or sample review so clients can assess your coding judgment and attention to detail. Applications are completed through the platform; if hired, work is usually delivered on a project-by-project basis with remote collaboration.
Expect projects to provide annotation guidelines, training examples, and a review flow. Your role may be hands-on labeling, designing the labeling schema, reviewing others’ annotations, or creating test suites for datasets. OpenTrain helps you find these opportunities, manage applications, and present your experience to hiring teams in the AI-training ecosystem.
Explore by specialty
More categories
By country
By language
