Research Scientist @ Kensho

Kensho is S&P Global’s hub for AI innovation and transformation. With expertise in Machine Learning and data discovery, we develop and deploy novel solutions for S&P Global and its customers worldwide. Our solutions help businesses harness the power of data and Artificial Intelligence to innovate and drive progress. Kensho’s solutions and research focus on speech recognition, entity linking, document extraction, automated database linking, text classification, natural language processing, and more.

Are you looking to solve hard problems and enjoy working with teammates with diverse perspectives? If so, we would love to help you excel here at Kensho. We are a collaborative group of experienced Research Scientists and Machine Learning Engineers, whose academic backgrounds include doctorate degrees in NLP, theoretical physics, statistics, etc. We take pride in our team-based, tightly-knit startup-like Kenshin community, which fosters continuous learning and a communicative environment.

At Kensho, we hire talented people and give them the freedom, support, and resources needed to accomplish our shared goals. We believe in flexibility-first and give our employees the opportunity to work from where they feel most productive and engaged (must be in the United States). We also value in-person collaboration, so there may be times when travel to one of our Kensho hubs (e.g., Cambridge, MA or NYC) will be required for team meetings or company events.

About the R&D Lab:

Since 2022, we have been building a world-class R&D lab comprised of NLP Research Scientists, and we heavily prioritize publishing in top-tier conferences. Our small team has demonstrated compelling results and is fueling innovation throughout Kensho and S&P Global at large. Specifically, we are continuously developing Large Language Models (LLMs) and are actively working on long-context question-answering (QA), complex reasoning, tokenization, alignment (e.g., factuality), multi-document QA, and more!

Our small team has reserved access to hundreds of fast GPUs (A100s), spanning Cloud and on-prem machines.

Our current projects include:

- Long-context document QA, where the answer is contained within documents that are hundreds of pages in length [1]

- Complex reasoning, including better understanding and improving models’ ability to approximate numbers (related to commonsense reasoning).

- Creating rigorous evaluation benchmarks, spanning domain knowledge, quantity extraction, and program synthesis [2]

- Improving existing alignment techniques for domain-specific needs, while also addressing factuality

- Dissecting tokenizers to better understand how each of the sub-components impact intrinsic and extrinsic performance [3][4]

- Multi-Document QA where the answer requires combining information from dozens of sources.

- Retrieval-augmented generation (RAG) methods

- Creating high-quality data filters for LLM development

Additionally, we maintain strong relationships with academia, including collaborating on several ongoing projects, providing industry grants, sponsoring conferences, and jointly holding faculty positions.

[1] DocFinQA: A Long-Context Financial Reasoning Dataset (Reddy et al., 2024)

[2] BizBench: A quantitative reasoning benchmark for business and finance (Koncel-Kedziorski et al., 2024)

[3] Tokenization Is More Than Compression (Schmidt et al., 2024)

[4] Greed is All You Need: An Evaluation of Tokenizer Inference Methods (Uzan et al., 2024)

Kensho states that the anticipated base salary range for the position is 150k-225k. In addition, this role is eligible for an annual incentive bonus and equity plans. At Kensho, it is not typical for an individual to be hired at or near the top of the range for their role and compensation decisions are dependent on the facts and circumstances of each case.

Technologies & Tools We Use:

ML: PyTorch, Weights & Biases, NetworkX
Deployment: Airflow, Docker, EC2, Kubernetes, AWS
Datastores: Postgres, Elasticsearch, S3

What You’ll Do:

Regularly reading late-breaking research papers and helping to identify pertinent directions of work
Developing novel, state-of-the-art NLP models that can scale to millions of documents
Working closely with other Research Scientists and ML Engineers
Writing clean, readable research code in PyTorch (not expected to write production-level code)
Contribute to a stellar engineering culture that values excellent design, documentation, testing, and code
Share your research results with your colleagues (presentations) and the world (published papers, patents, and blog posts)

What You’ll Need:

Outstanding people come from all different backgrounds, and we’re always interested in meeting talented people! Therefore, we do not require any particular credential or experience. If our work seems exciting to you, and you feel that you could excel in this position, we’d love to hear from you. That said, most of our successful candidates possess the following, which reflects both our technical needs and team culture:
Hold a PhD in Computer Science or related field (or a Master’s with significant research experience)
Have published in a top-tier ML/NLP conference (e.g., ACL, NAACL, EMNLP, NeurIPS, ICML)
Are proficient in writing code in PyTorch, Tensorflow, or JAX
Have experience with the techniques required to work effectively with large, messy real-world data
Prefer to collaborate iteratively on hard problems with your teammates rather than spending stretches of time working alone and presenting your results intermittently
Have a love for learning new skills and domains
Are excited to share knowledge freely, proactively, and effectively with others who are interested
Are a generous teammate who takes work seriously without taking yourself too seriously

At Kensho, we pride ourselves on providing top-of-market benefits, including:

- Medical, Dental, and Vision insurance

- 100% company paid premiums

- Unlimited Paid Time Off

- 26 weeks of 100% paid Parental Leave (paternity and maternity)

- 401(k) plan with 6% employer matching

- Generous company matching on donations to non-profit charities

- Up to $20,000 tuition assistance toward degree programs, plus up to $4,000/year for ongoing professional education such as industry conferences

- Plentiful snacks, drinks, and regularly catered lunches

- Dog-friendly office (CAM office)

- Bike sharing program memberships

- Compassion leave and elder care leave

- Mentoring and additional learning opportunities

- Opportunity to expand professional network and participate in conferences and events

About Kensho

Kensho is an Artificial Intelligence company that builds solutions to uncover insights in messy and unstructured data that enable critical workflows and empower businesses to make decisions with conviction. We were founded in 2013 and now serve as S&P Global’s innovation hub. We continue to maintain our distinct, independent brand in order to best promote our breakthrough, innovative culture. Our team of Kenshins enjoy a dynamic and collaborative work environment that runs autonomously from S&P Global, while leveraging the unparalleled breadth and depth of data and resources available as part of S&P Global. As Kenshins, we pride ourselves on maintaining an innovative culture that depends on diversity and inclusion.

We are an equal opportunity employer that welcomes future Kenshins with all experiences and perspectives. Kensho is headquartered in Cambridge, MA, with an office in New York City. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, or national origin.

Research Scientist

NLP

Kensho

Compensation Overview

Technologies & Tools We Use:

What You’ll Do:

What You’ll Need:

Simplify's Take

What believers are saying

What critics are saying

What makes Kensho unique

Benefits

Company News