AI Product Operator, Niural AI Labs ( Night Shift )
Hybrid (US (night shift))full-timePosted 4w ago
About the role
Research-centered role inside Niural's product team to design and build the Learning Gym for training and evaluating AI agents on real Niural workflows (e.g., payroll, contractor onboarding, invoice reconciliation, compliance). Combine environment engineering, reward design, RL training, and first-author papers, with production-quality delivery and night-shift US hours.
Responsibilities
- Build sandboxed environments that wrap real Niural workflows, stateful across episodes so actions persist across tasks
- Seeded and replayable environments to reproducibly reproduce results
- Implement difficulty knobs and curriculum structures for edge-of-ability environments
- Incorporate realistic data issues (missing fields, stale records, contradictory sources) using payroll data
- Generate synthetic data for environments while complying with data governance
- Design programmatic verifiers from ground truth, minimize reliance on model judging other models
- Score trajectories and support multiple correct paths
- Build against reward hacking and data contamination
- Run reinforcement learning loops: training, curriculum schedules, fine-tuning, distillation
- Establish baselines, run ablations, test generalization, publish negative results internally
- Prototype fast with 1–2 week spikes on agent architectures and tools
- Hand off successful approaches to production teams and stay involved during deployment
- Write first-author research papers with reproducible methodology and honest limitations
- Produce internal technical reports and represent work externally via preprints, talks, and open sources
Qualifications
- A published, peer-reviewed paper at a conference or journal with clear attribution
- Foundational knowledge of AI and machine learning, including understanding training processes
- Strong Python skills with numpy and scientific stack, readable to others
- Real engineering skills: building and running services (APIs, databases, queues, Docker, deployment, logging)
- Experience with agent systems, harnesses, tool use, and multi-step workflows
- Reinforcement learning literacy with ability to implement policy optimization and reason about reward design
- Reproducibility as a habit (cloning repos, clear READMEs, runnable code)
- Evidence of impactful work (products built, cited results, open-source usage, or decision-changing impact)
- Willingness to work permanent night shift on US hours
Skills
Required:PythonMachine Learning / Reinforcement LearningSystems / DevOpsResearch & Writing
Benefits
- Competitive salary package
- Access to continuous learning and career advancement programs
- Mentorship and coaching
- Collaborative work environment
- Complimentary snacks and lunch
- Latest MacBook Pro and high-performance monitor
- Fast-track university admission opportunities for CS Masters in the US
- Visa sponsorship for high performers (L1 and Green Card)
Last updated Oct 3, 2026. Interested? Apply on the company site.
About Niural
One HR & Payroll Platform. From First Hire to IPO.
Niural is the future of HR management with centralized control, advanced analytics, and automation for streamlined HR operations worldwide.
HRPayrollPEOEORglobal payrollcontractor managementNiural PayAI
More open roles at Niural
- Marketing Web Developer2d agoMarketing
- Sales Development Representative (SDR)1w agoRemote · Sales
- Manual QA2w agoRemote
- Customer Support Specialist2w agoHybrid (Remote)
- Graphic Designer2w agoHybrid (Remote)