Insurance underwriting isn’t for the weak. It’s a dizzyingly complex undertaking that requires connecting data points across disparate sources to support consequential decisions—all while meeting modern expectations for speed, accuracy, and compliance. 

The role has grown exponentially more challenging as technology has become more ubiquitous, stretching our information-rich digital trails ever longer.

Over the past two decades, various vendors have developed Intelligent Data Processing (IDP) tools to manage all this information by automating the extraction, ingestion, and structuring of data at scale. These tools have been widely adopted by carriers, but fall short of today’s mounting data challenges–in fact, they’re exasperating them.

McKinsey estimates that underwriters spend 30-to-40% of their time on rote administrative tasks “such as rekeying data or manually executing analysis.” These were the types of tasks that IDPs were supposed to automate and make more efficient—but that’s not what’s happening. In a recent Accenture survey, 64% of underwriters reported that today’s tech either makes no difference or increases their workload. 

Automated data extraction was, until recently, the only way to tame the information deluge. New technologies have paved the way for a better, more seamless approach. Emerging LLM-powered AI represents a new paradigm that eliminates extraction chokepoints, reduces the burden on overtaxed underwriters, and accelerates decisioning.

Generative AI in insurance changes everything

Traditional IDPs were designed to exhaustively extract every piece of data–no matter how irrelevant or repetitive—so that it can be structured into a centralized database and passed along to overloaded human underwriters to query and scrutinize. The more complex and document-laden a process (e.g., loss run reports with intricate hierarchical ordering of nested sets), the more odious the inefficiencies and the more work tossed onto underwriters’ plates.

Insurance solutions touting the “most efficient” or “fastest” data extraction are about as meaningful in 2023 as boasting the “highest print-quality” fax machine. Comprehensive extraction is a relic of a fading technological paradigm. The industry is rightly turning to next-gen AI technologies to free underwriters from repetitive data work (which is better handled by machines anyway) so they can focus on building value and closing deals. 

Sixfold uses state-of-the-art LLMs to synthesize information across multiple sources and generate summaries in plain language for underwriter review. No processing power is misspent on redundant extraction; underwriters’ valuable time is no longer wasted sorting through virtual buckets of well-structured (but context-free) data. 

When processing a life insurance application, traditional IDPs will, for example, extract each mention of the applicant having diabetes, even if it appears across dozens of documents. Unlike AI-powered platforms, IDPs are incapable of discerning meaning from data—underwriters are still required to connect the dots. Sixfold skips the needless chronicling of data points and independently generates clear summations of relevant throughlines (e.g., “The applicant was diagnosed with type 2 diabetes 12 years ago and it’s being properly managed with insulin and diet”), thus freeing underwriters to forgo the data work and render decisions faster.

Sixfold brings the power of advanced AI to underwriting

In effect, Sixfold provides underwriters with a virtual army of researchers, data processors, and writers who know precisely what information is needed to render decisions quickly (and just as importantly, what isn’t). 

It’s already having a huge impact. With Sixfold, companies are accelerating submission-to-quote cycles by as much as 43%, clearing backlogged queues, and massively increasing GWP per underwriter. 

Even better? It’s far easier to get up and running with Sixfold than a traditional IDP. These older systems required huge investments in time and resources to train their ML models on an organization’s unique needs. Sixfold, on the other hand, can be easily—and quickly—configured to match the appetite and needs of specific carriers and programs. It’s more-or-less ready to go out-of-the-box (or out of the virtual SaaS box).

AI is reshaping insurance before our eyes

The marketplace is littered with the remnants of corporate behemoths that misread the technological tea leaves—and in today’s world, giants fall fast. Consider how, in just one decade, Yahoo slid from the world’s most popular website to near-irrelevance. Or how Kodak only took eight years to complete its journey from top-five global brand to ejection from the Dow Jones. Or how, in a mere six years, Blockbuster leaped from its 9,000-plus-location peak into bankruptcy. 

The takeaway: Past performance will not save you. New technological paradigms can seemingly come out of nowhere to reward leaders who had an eye on the future—and expose those who didn’t.

I’m confident that 2023 will be remembered as an inflection point for generative AI. The way insurance is handled moving forward will be a radical departure from the past. There’s now a clear industry-wide divide between those pursuing iteration and those seeking transformation. Which side do you want to be on?

We recently sat down for a quick Q&A with Stewart Hu, AI scientist at Sixfold. Our conversation ranged from his career journey to how he stays current in the field, as well as the tasks on his daily agenda.

Let’s get this started! In your own words, what does your job as an AI scientist involve?

AI scientists engage in a lot of practical work. Despite our 'scientist' title, our roles often overlap with those of developers or research engineers. In fact, over 50% of our tasks are typical software engineering activities. We develop software grounded in foundational models, employing a range of techniques, not just AI.

Previously, AI encompassed anything linked to machine learning, but now it's more commonly associated with large language models like GPT. Our role includes integrating these models into software applications, utilizing models such as GPT-4, and even fine-tuning our custom models. Additionally, we apply both traditional machine learning and deep learning methods. This involves creating classifiers with techniques predating neural networks, like gradient boosting machines or random forests. At our core, we are software engineers crafting machine learning algorithms to address real-world challenges.

How did you get into the world of Generative AI? 

My fascination with AI really took off with GPT-3's emergence. But it was the debut of the stable diffusion model in August 2022 that truly captivated me. This revelation prompted me to pivot my career towards a tech startup specializing in deep learning and AI.

In the early stages of my career, I worked as a software engineer. This was followed by a ten-year journey in data science, beginning with statistical learning and gradually evolving into machine learning, deep learning, and finally AI. Essentially, I devoted my first decade to hardcore software development, and the next decade explored the realms of data science and machine learning.

Could you give some insights into what's on a typical to-do list for you?

My work is basically divided into three key areas.

Firstly, there's data management: sourcing appropriate data, organizing it properly, and conducting thorough analyses. A major chunk of our time is dedicated to dealing with data - acquiring, scrutinizing, and delving into it.

Secondly, I engage in software development, where my goal is to craft software that's not only reusable but also adaptable to growing complexities. This involves strategic software design to ensure it can be easily scaled up.

The third area is AI, particularly focusing on 'retrieval augmented generation’ . This entails extracting pertinent details from extensive document collections to accurately contextualize models like GPT-4. My day-to-day involves juggling these three components.

How would you distinguish a purpose-built AI tool from a generic one?

AI often gets hyped up with flashy demos requiring little coding. However, Sixfold is a purpose-built Gen AI tool, our focus is on crafting solutions that address real-world business problems, not just making eye-catching demos. We use AI to make underwriters work faster, more accurate, and enjoyable. By taking over repetitive tasks, AI allows underwriters to focus on the more engaging aspects of their job.

Our platform is built with a strong emphasis on accountability, not just on interpretability or explainability. This means our solutions cite sources when making recommendations and provide actual source documents for our classifications. It's a practical, business-centric approach that boosts confidence in underwriting decisions.

What excited you the most about joining Sixfold?

Two things particularly drew me to Sixfold. First, the experienced team leading the company. The founders have a proven track record of creating substantial business value, blending tech knowledge with sharp business insight. Second, on a personal level, my wife has been in the insurance industry for over ten years, and I've always found it fascinating. Joining Sixfold presented a chance to dive deeper into this sector. 

It was the combination of the seasoned leadership and the company's expertise in insurance and underwriting that ultimately convinced me to become part of the team.

How do you stay engaged with the AI community? 

My go-to resource is X (formerly known as Twitter), where I've created a list named ‘AI Signals.' This list features over 100 experts deeply engaged in the field, tackling everything from fine-tuning models to enhancing the speed of large language model inference. While some of these individuals may not be widely known, their insights are incredibly valuable. 

Previously, I would follow arXiv for academic papers, GitHub for trending repositories, and Papers with Code to find research papers with their corresponding code. However, X has become my most essential tool. I regularly check updates from my list there to keep up-to-date with the latest developments.

That sounds like a great list! Can we share it with the readers?‍

Of course, happy to share it - here you go! 

How can people best follow your work?

I haven't been active on my blog lately, but I do maintain a GitHub repository named 'LLM Notes.' It serves as a practical guide for data scientists and machine learning practitioners. This repository is a compilation of the knowledge and insights I've gathered throughout my career. A few months back, I uploaded a wealth of information there, including lessons learned, common pitfalls, and personal experiences. It's a good resource for anyone interested in the field. 

Thanks for your time, Stewart! We’ll let you get back to your to-do list now.

‍Interested in joining the Sixfold team? Take a 👀 at our open positions here.

AI is the defining technology of 2023. After years of unfulfilled promises from Hollywood and comic books, the science fiction AI future we’ve long been promised has finally become business reality. 

We can already see AI following a familiar path through the marketplace similar to past disruptive technologies. Stage one: it’s embraced by early adopters before the general public even knows it exists; stage two: cutting-edge startups tap these technologies to overcome long-standing business challenges; and then stage three: regulators draft rules to guide its usage and mitigate negative impacts.

There should be no doubt that AI-powered insurtech has accelerated through the first two stages in near record time and is now entering stage three. The Colorado Department of Regulatory Agencies recently adopted regulations on AI applications and governance in life insurance. To be clear, Colorado isn’t an outlier, it’s a pioneer. Other states are following suit and crafting their own AI regulations, with federal-level AI rules beginning to take shape as well.

AI underwriting solutions, meet the rule-makers

The early days of the regulatory phase can be tricky for businesses. Insurers are excited to adopt advanced AI into their underwriting tech stack, but wary of investing in platforms knowing that future rules may impact those investments. 

We at Sixfold are very cognizant of this dichotomy: The ambition to innovate ahead, combined with the trepidation of going too far down the wrong path. That’s why we designed our platform in anticipation of these emerging rules. 

We’ve met with state-level regulators on numerous occasions over the past year to understand their concerns and thought processes. These engagements have been invaluable for all parties as their input played a major role in guiding our platform’s development, while our technical insights influenced the formation of these emerging rules.

To simplify a very complex motion: regulators are concerned with bias in algorithms. There’s a tacit understanding that humans have inherent biases, which may be reflected in algorithms and applied at scale.

Most regulators we’ve engaged with agree that these very legitimate concerns about bias aren’t a reason to prohibit or even severely restrain AI, which brings enormous positives like accelerated underwriting cycles, reduced overhead, and increased objectivity–all of which ultimately benefit consumers. However, for AI to work for everyone, it must be partnered with transparency, traceability, and privacy. This is a message we at Sixfold have taken to heart.

In AI, it’s all about transparency

The past decade saw a plethora of algorithmic underwriting solutions with varying degrees of capabilities. Too often, these tools are “black boxes” that leave underwriters, brokers, and carriers unable to explain how decisions were arrived at. Opaque decision-making no longer meets the expectations of today’s consumers—or of regulators. That’s why we designed Sixfold with transparency at its core.

Customers accept automation as part of the modern digital landscape, but that acceptance comes with expectations. Our platform automatically surfaces relevant data points impacting its recommendations and presents them to underwriters via AI-generated plain-language summarizations, while carefully controlling for “hallucinations.” It provides full traceability of all inputs, as well as a full lineage of changes to the UW model, so carriers can explain why results diverged over time. These baked-in layers of transparency allow carriers–and the regulators overseeing them–to identify and mitigate incidental biases seeping into UW models.

Beyond prioritizing transparency, we‘ve designed a platform that elevates data security and privacy. All Sixfold customers operate within isolated, single-tenant environments, and end-user data is never persisted in the LLM-powered Gen AI layer so information remains protected and secure.  

Even with platform features built in anticipation of external regulations, we understand that some internal compliance teams are cautious about integrating gen AI, a relatively new concept, into their tech stack. To help your internal stakeholders get there, Sixfold can be implemented with robust internal auditability and appropriate levels of human-in-the-loop-ness to ensure that every team is comfortable on the new technological frontier. (Want to learn more about how Sixfold works? Get in touch.)

We at Sixfold believe regulators play a vital role in the marketplace by setting ground rules that protect consumers. As we see it, it’s not the technologist’s place to oppose or confront regulators; it’s to work together to ensure that technology works for everyone.