Top Tech Jobs & Startup Jobs in Austin, TX

Build and validate challenging Codebase Q&A tasks for AI agents using production Go repositories. Responsibilities include exploring unfamiliar codebases, understanding runtime behavior and history, rewriting engineering questions, refining rubrics and foils, running validation workflows, documenting tasks, and responding to expert review. Work is fully remote, asynchronous, and paid per approved task.

Evaluate AI-generated financial plans, budgets, forecasts, workflows, and reporting artifacts for accuracy, financial soundness, regulatory compliance, and analytical rigor. Compare AI responses and provide expert feedback on cash flow oversight, budgeting, pricing, and financial reporting. Requires experienced financial management expertise and a commitment of at least 20 hours per week for an approximately four-to-six-week project.

Design reinsurance and program-business scenarios, assess AI-generated responses, create market-quality reference materials, and provide detailed feedback on treaty structures, delegated authority, fronting, captives, risk transfer, contract logic, financial responsibilities, and market realism.

Capture and evaluate natural Japanese voice interactions with an AI assistant using a Mac. Complete everyday productivity tasks by voice, transcribe and clean spoken queries, rate AI responses, record task data in structured sheets, and upload supporting recordings. Candidates must be native or near-native Japanese speakers, understand English instructions, own an Apple Silicon Mac running macOS Sequoia 15.0 or newer, and be comfortable with productivity applications and detailed quality review.

Capture and evaluate natural Thai voice interactions with an AI assistant using a personal Mac. Complete everyday tasks through voice commands, transcribe and clean spoken queries, rate AI responses, maintain structured task logs, and upload recordings. Applicants must be native or near-native Thai speakers, understand English instructions, use common productivity applications, and own an Apple Silicon Mac running macOS Sequoia 15.0 or newer. The project lasts four weeks and begins immediately.

Annotate Korean PDF pages by identifying, bounding, typing, and ordering document regions; link related components; transcribe Hangul, hanja, and handwriting exactly; capture page metadata; and flag illegible or unsuitable pages. Experienced annotators also review colleagues’ work. The role requires exceptional character-level accuracy, consistent taxonomy application, and comfort with complex layouts such as tables, diagrams, newspapers, exams, and handwritten forms.

Design complex insurance claim and coverage scenarios for AI model training, including disputed coverage, litigation, allocation, bad-faith risk, and major-loss settlement strategy. Produce senior-level claim analyses, authority requests, reserve rationales, and litigation plans; evaluate AI responses against claims standards; identify unsupported or improper positions; and provide written feedback. Participate in onboarding and calibration sessions.

Design and evaluate life insurance and annuity advisory, underwriting, policy-service, suitability, and replacement scenarios for AI training. Create expert reference responses, assess model outputs for product accuracy, mathematical correctness, suitability, tax treatment, and disclosure compliance, and provide feedback to improve model performance. Participate in onboarding and calibration sessions.

Build and validate reinforcement-learning tasks for AI agents exploring unfamiliar C repositories. Responsibilities include investigating runtime behavior, rewriting engineering questions, refining rubrics and foils, running validation workflows, documenting solutions, and addressing expert review. Work focuses on production systems such as PostgreSQL, Redis, OpenSSH, OpenSSL, and other foundational C codebases.

Annotate Japanese PDF pages by identifying, bounding, typing, and ordering document regions; link related components; transcribe all text exactly, including handwritten and vertically arranged content; capture metadata; and flag illegible or unsuitable pages. Experienced annotators also review colleagues’ work. Native Japanese fluency and command of kanji, hiragana, katakana, furigana, and variant characters are essential.