Applied AI Researcher
AI and Research · California or Remote within approved U.S. states · Full-time · Senior
Improve the quality, reliability, and usefulness of AI-supported product experiences — applied evaluation, retrieval, product assistance, workflow recommendations, knowledge quality, safety controls, and measurable improvements rather than unrestricted foundational-model research.
About SendiMessage
SendiMessage builds managed messaging infrastructure: dedicated messaging lines, one clean API for SMS and supported iMessage delivery, delivery tracking, incoming replies, and signed webhooks. We work carefully — controlled launches, evidence-based decisions, honest claims — and we hire people who like working that way.
What you will do
- Design evaluation datasets for product-assistance systems
- Evaluate answer accuracy and unsupported-claim risk
- Improve retrieval and knowledge-ranking quality
- Research conversational product experiences
- Analyze failure cases and repeated user confusion
- Develop measurable AI quality benchmarks
- Prototype AI-supported operational workflows
- Build evaluation and regression-test systems
- Improve grounding and citation behavior
- Develop guardrails for sensitive or uncertain answers
- Collaborate with product and engineering
- Translate research findings into production recommendations
- Document assumptions, experiments, and results
What we are looking for
- Professional or academic experience in applied AI, machine learning, NLP, information retrieval, or AI evaluation
- Strong understanding of large language model behavior
- Experience with evaluation methodology
- Ability to analyze hallucination, retrieval, and grounding failures
- Strong written communication
- Ability to work with engineers and product managers
- Experience using Python and relevant AI tooling
- Evidence-based research approach
Preferred experience
- Retrieval-augmented generation
- LLM evaluation
- Search and ranking
- Conversation systems
- AI safety or guardrails
- Agent evaluation
- Human-in-the-loop systems
- Production AI monitoring
- Developer tools
What success looks like
- Assistant answers become measurably more accurate and honest
- Failure modes are catalogued with mitigations
- Evaluation runs on every knowledge change
Location and work arrangement
California or Remote within approved U.S. states. This role does not offer employment sponsorship at this time. Remote work is within approved U.S. states — confirmed during the hiring process.
Compensation and benefits
- Base salary:
- $140,000–$190,000 per year
The expected base salary range for this role is $140,000–$190,000 per year. Final compensation will depend on relevant research experience, applied AI expertise, technical depth, location, and employment structure.
Benefits will be discussed during the hiring process based on employment arrangement and location.
Apply for this role
SendiMessage evaluates candidates based on role-related qualifications.