Online, self-serve life insurance company.
United States of America
At Ethos, we strive to provide every potential customer with the the right insurance for each specific situation. We achieve this by making informed product selection and underwriting risk decisions based on data from and about the customer. How do we best generate, collect, and consume this data and how does this uniquely position us to win? You will be the author and owner of this strategy - you will develop a plan for how the data we generate will be used to improve the product over time and why this creates a defensible moat to increase the product and company's chances of long-term success.
Core areas of ownership
Underwriting data platform
As we grow and collect more underwriting data, we need frameworks to rapidly enable powerful models and engines. The use cases in these workflows are tightly integrated with the product so an off the shelf or open source solution won’t work. In effect, by owning this, you’ll aim to abstract away the need to manage data, and deploy and monitor results; you’ll then focus on iterating on the model and the experiments themselves.
Standardization of underwriting data
There is no central place to see all the data that is used to determine pricing and personalization, or how it connects to the rest of the data Ethos has. Without a structure to centralize and document metadata, analyzing our products and underwriting becomes an exercise in comparing disparate entities and customer experiences. It becomes unclear what the data actually means, where they come from, how reliable they are, etc. You will resolve this by standardizing in a way that unlocks almost infinite possibilities of how the data can be used.
Products and features such as pricing and product recommendations, underwriting and risk assessment, fraud detection, sales and underwriting optimization, etc. lend themselves naturally to ML / AI solutions. You will own thinking about how data can be leveraged to improve processes and products (e.g. analyze customer journeys to predict and optimize the underwriting of future customers) or how to design an entirely new experience altogether (product matching for customers of different cohorts/segments). Ultimately, your work will go towards directly improving key business metrics for user-facing products and features.
Questions you’ll answer/tackle:
- What’s the optimal product for a given applicant using attributes/demographic characteristics we learn about the applicant via self disclosed information, as well as on and off site behavior and other contextual clues/data sources?
- How might we develop new underwriting rules and ways to underwrite efficiently? What does it mean to underwrite efficiently?
- What data should we be using that we don't have today? How would we incorporate it, what would its impact be, and how do we get it?
- How will a model’s success be evaluated over time? How can we drive improvements in process and pricing, specifically as it pertains to rules making and underwriting?
- Do our ML models need to be scored in real-time or can they be pre-scored offline?
- How do we detect if someone is lying based on the time they spend on a page? How do we have the best in class audit and fraud monitoring systems?
- What is the plan for retraining models on new data?
- What kind of infrastructure is needed to support our current and future product(s)?
- What is the complexity cost for implementing a model in production? Other stakeholders will offer input when answering these questions but as the Product Lead, you will weigh the tradeoffs involved in product development and own the final decision.
What you’ll be expected to know and/or learn
- Stakeholders you will work closely with include those on the following teams: Data, Engineering, Risk, Underwriting, Business Analytics, Product Analytics, CX/CS, Sales, and Partnerships.
- You will be expected to learn (or already know) that collecting data and using data are two different parts of building with data with different tradeoffs and often involving different parts of the engineering team. And you will help drive the product development process so that the handoff between these two processes is seamless.
- You will translate requirements between engineers, designers, marketers, and other PMs, building product instrumentation and data storage into your product acceptance criteria while collaborating with data scientists/analysts to ensure that data will be accessible and usable for analysis and modeling as soon as possible. You won’t leave it to engineers who are not data scientists to make assumptions about what kinds of data will be valuable for others.
- Data, models, and outputs aren’t enough; you should ladder these components back up to the business model and Ethos’ strategy. You must have a very strong understanding of our strategy to think proactively and plan appropriately. Leveraging data in a way that doesn’t align with our strategy will not only waste time and money for no reason, it will hinder our ability to meaningfully improve other areas of the business.
- You will need to develop a deep understanding of data modeling, data infrastructure, and statistical and machine learning, well beyond understanding the results of experiments and reading dashboards. You will be the source of truth for questions regarding what is possible and what will soon be possible by taking full advantage of the flow of data. Traditional PMs operate at the intersection of business, engineering, and user experience; you will operate at this intersection but must also have domain knowledge of data and data science.
- 8+ years of tech or finance experience
- 5+ years experience in developing quantitative products including experience with at least two of the following: ML, AI, risk assessment and detection, pricing, and sentiment analysis
- Exceptional organizational and analytical experience
- Expert with SQL
- Bonus points for experience in pricing, behavioral economics, and operations research
Everyone is welcome at Ethos. We are an equal opportunity employer who values diversity and inclusion and look for applicants who understand, embrace and thrive in a multicultural world. We do not discriminate on the basis of race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, or disability status. Pursuant to the SF Fair Chance Ordinance, we will consider employment for qualified applicants with arrests and conviction records.