Applied Scientist (ML), MTV

Role

As an Applied Scientist (ML) at Samaya, you will collaborate closely with our product and engineering teams, and use cutting-edge ML research to transform how users interact with Samaya in their daily workflows. You'll drive impact across the entire ML lifecycle: from problem formulation and system analysis to data collection, benchmark development, model training, and production deployment. You’ll also have opportunities to publish your work and deliver impact to the ML community. Your expertise will advance our capabilities in these critical technical domains:

• Retrieval, ranking and RAG
• LLM post-training and reinforcement learning
• AI agents for knowledge workflows
• ML benchmarks

Your work has the potential to transform the following key Samaya products and deliver impacts to tens of thousands of professional users:

Instant QA: Our custom-built Question Answer system using state of the art in house models, seamlessly trained to work together to enable instant expert intelligence.

Agents: You will enable expert-level agentic workflows to automate comprehensive knowledge work and enable AI tools that work with experts to gain new insights.

You can read some of our previous ML work at: https://samaya.ai/blog/ and https://samaya.ai/research/.

Responsibilities

• Formulate an ML problem from product requirements.
• Analyze existing ML systems for their limitations, and propose and validate novel methodologies to improve upon existing systems.
• Create and productionize cutting-edge research prototypes for knowledge work at scale.
• Build novel ML evaluation datasets that serve as crucial criteria for production feature rollouts.
• [Optionally] Mentor ML interns and publish your research findings with the community.

Experience

Required

• PhD or Master’s degree in Computer Science, Machine Learning, NLP, or a related field.
• Strong background in deep learning, large language models, and NLP techniques.
• A strong track record of first-author publications in top AI/NLP conferences (e.g., NeurIPS, ICML, ACL, EMNLP).
• Proficiency in Python and deep learning frameworks such as PyTorch or Transformers, and strong coding skills.

Preferred

• 2+ years of experience in an industry applied ML research environment
• Familiarity with retrieval-augmented generation, reasoning, LLM training and reinforcement learning techniques.

Compensation

The cash compensation range for this role is $190,000 - $275,000.

Final offer amounts are determined by multiple factors, including experience and expertise, and may vary from the amounts listed above.

In addition to the base salary, we may consider equity as part of our total compensation package.

Benefits

• Comprehensive health insurance coverage (medical, dental, vision, and short-term disability) to support your health and wellbeing
• Support your long-term financial well-being with 401K (US) and enhanced Pension contributions (UK)
• Flexibility to rest and recharge with unlimited PTO
• Travel budget to provide opportunities for learning and collaboration by attending conferences, visitings other offices, and more
• Office equipment allowance to enhance the comfort of your workspace
• Hybrid work environment to promote balance and flexibility, with typically 3 (or more) days in the office per week

Inclusive Hiring

Interview Accommodations: We are committed to ensuring an equitable selection process for everyone and welcome applicants from varied backgrounds to enrich our team. If you require accommodations or adjustments during our recruitment process, please inform us.

Equal Opportunity Employer: We do not discriminate on the basis of race, color, religion, sex (including pregnancy and gender identity), national origin, political affiliation, sexual orientation, marital status, disability, genetic information, age, membership in an employee organization, retaliation, parental status, military service, or other non-merit factor.

Visa Sponsorship: We do sponsor visas! However, we aren't able to successfully sponsor visas for every role and every candidate. If we make you an offer, we will make every reasonable effort to get you a visa, and we retain an immigration lawyer to help with this.

About Samaya

Samaya is building the first Human-AI Knowledge Network – an information ecosystem to transform expert knowledge work.

Expert knowledge work drives trillions of dollars of economic activity. Teams of human experts painstakingly synthesize insights and drive decisions from vast volumes of noisy, real-time information. For examples, in financial services, expert analysts hone in on and use key economic insights to inform high-stake investment decisions. Although the volume of information only continues to grow, past technology has only had a passive role, and today's human experts struggle with information overload. We believe AI should take an active role in complex knowledge work, becoming an equal collaborator to human experts.

But such an AI will not "simply emerge" through scaling, or the development of general purpose LLMs. At Samaya, we are developing an AI-system purpose-built for "expert intelligence" -- for reasoning and interacting with real-world information networks. Our AI is designed to consume dense, noisy real-time information, distill key insights, form connections, contextualize findings, and make expert predictions. We are building a future where our expert intelligence AI can transform global knowledge work for the better.

Our Operating Principles

• Put Users first. Our users rely on us to do their jobs. We exist because our users trust us to help them achieve their goals. In return for this trust users place in us, we keep their needs as our top priority.
• Win as a collective. We are high achievers with a drive to succeed. We build strong bonds over this shared drive. We dive in to help when one of us needs it. We’re kind to each other and boost each other to succeed and grow professionally and personally. We build trust with each other by making commitments and consistently delivering on them. This trust means we genuinely support each other, embracing feedback as a tool for growth and improvement. We win by operating this way, as one team.
• Focus and iterate quickly. Bias for action makes us build and learn quickly. Iterating fast requires clarity on what outcomes we are targeting and why. Prioritizing the important things, taking full ownership and initiative, making fast initial progress, and rapid iterations lead to the best outcomes.
• Innovate Relentlessly. We pursue novel insights, challenging the status quo and reimagining how things are done. We aren’t attached to the past when improving our product and how we work in the future. We actively invest time in innovation, thinking “outside the box” to consistently raise our standards.
• Prioritize Outcomes over Egos. We are committed not to a person, an idea, or an opinion but to continuously making progress to our goals. Sometimes, our goals are ambiguous; in those moments, we iterate, learn, and move on to the next inquiry. We ask the tough questions with kindness, dropping our egos in our pursuit of evidence. For our business goals, we learn from our users. For our scientific goals, our understanding is built through rigorous experimentation, research, and observation. For our personal goals, we embrace candid feedback and collaborative learning to guide our progress.