EliseAI

The challenge: Keeping up with the demand for data, in the midst of lightning growth

Healthcare and housing run on paperwork, but EliseAI is changing that — replacing these slow, manual processes with AI that works. The company has doubled in size every year for four years running, but with that growth came a new challenge: a longer ticket queue for the data team.

The data team was moving fast, but the business was moving faster. Their existing BI and analytics tool setup straddled between being too rigid and too technical for their team’s needs — and as demand for answers accelerated across every team, it became clear they needed something that could serve data science and business users on one platform. The team searched for a modern and flexible tool capable of bespoke data apps, ad hoc analytics, and giving users the answers they needed regardless of technical skill.

The solution: Getting trusted data in the hands of every rep with Snowflake and Hex

EliseAI's data lives in Snowflake, processed and modeled through dbt. Hex became the analytics layer on top — connecting directly to Snowflake and opening that data up to everyone, regardless of technical skill.

The impact showed up quickly. In the past year, active users tripled as the company grew with 91% of the company pulling or accessing data in Hex on a regular basis. The data was always there. Now, for the first time, the whole company could reach it.

Hex not only gave them the flexibility to build custom data solutions, it also enabled the whole org to move faster. Now, analysts could instruct Hex’s Notebook Agent in natural language and watch it generate SQL, Python, charts, and entire data apps. Business users could use Threads, Hex's conversational interface, to get trustworthy answers without waiting on anyone or writing a query.

Hex changed how EliseAI's entire org interacts with data. Over the course of the past year, usage of Hex skyrocketed across 14 teams, who built or viewed one of EliseAI’s 7.3K projects.

Reducing QBR prep from 2 days to 5 minutes, with a data app

When a customer signs their contract with EliseAI, their post-sales team gets to work — guiding clients through onboarding and holding quarterly business reviews (QBR). For the team, a good QBR was a strategic moment: a chance to highlight wins, maximize platform usage, and identify where an account can grow. But for a long time, getting ready for one was anything but strategic.

Previously, a non-technical team member could only access pre-built dashboards or data available in existing databases. Building a QBR deck for each customer with custom insights and charts took two days, but when the EliseAI team onboarded Hex, Hannah Lee, an Analytics Engineer on the data team, was able to automate the entire process with a single Hex app. Now, with just three clicks, reps could get a full slide deck, with data merged into insights from Snowflake, Salesforce, and product databases. Reps now walk into QBR calls with slides of performance trends, portfolio details, and active integrations. And, if a conversation went deeper, they could now pull from a massive bank of slides that were pre-populated with various data cuts to tailor the narrative. Even better, analysts could trace every data point back to its source via the notebook view.

This depth of preparation has led to a marked step-change in the quality of customer conversations and opened the door for business expansion.

"Our customer success teams are having more strategic conversations now because of Hex. With increased access to data, our team is now leading customer calls with detailed insights, and seeing a clearer picture of performance." — Hannah Lee, Analytics Engineer

What used to be a two-day slog is now a five-minute workflow that every customer-facing team at EliseAI relies on — SDRs, sales, and post-sales. And with the Notebook Agent, building the app took just one day, instead of weeks, leading to a 97% reduction in development time.

A 98% reduction in audit prep and a clearer path to growth

A strong QBR doesn't just show customers how they're performing, it also highlights room for better product usage. When clients see their numbers laid out clearly, the next question is usually: what could we be doing differently to improve them?

It's a deceptively simple ask, but answering it used to require at least two hours of work: checking which features a customer had enabled, copying that usage into a spreadsheet, and hoping nothing had changed by the time the meeting started.

To solve this, Strategy and Operations Lead, Ian D'Silva, used Hex’s Notebook Agent to build an app that handled customer audit prep automatically. The agent modeled the architecture from the existing QBR data app that Hannah built, without involvement from any analytics engineers. The full app, which could’ve taken weeks to build, now took just a couple of days, resulting in an 80% time savings using the Notebook Agent to create an MVP.

The app generates a full customer audit deck in minutes. The app scans a customer account and flags which features they are and aren’t using. It then calculates the performance lift that each unused feature would unlock. From there it generates a ready-to-present deck unpacking the information, and automatically links each feature to EliseAI’s live product settings, so reps can click through and make changes live on the call.

Threads: Turning one-off questions into instant answers

Beyond the bigger workflows, Ian and others also use Threads — Hex’s conversational data interface — for one-off questions that used to fall through the cracks. Threads queries Snowflake directly, so answers are trusted, without anyone writing SQL.

"Hex helped us lower the barriers to accessing data, which made everyone’s workflows more data-driven.” — Ian D'Silva, Strategy and Operations Lead

Before Threads, figuring out "who hasn't had a QBR in the last 90 days?" required Ian to either go to a spreadsheet that might be out-of-date, write a query, or build a dashboard for a one-off question. Now he’s able to type the question in natural language, let Hex run in the background, and receive an answer within a few minutes.

Others on the team are also experiencing the value of Threads. One of Ian’s favorite use cases of Threads is when customers ask account reps a question on the call, reps can easily use natural language to get real-time insights and have an answer for the customer before hanging up. This reduces the amount of follow-up or time spent querying to get answers to each customers’ nuanced question.

"Hex’s Threads allows a lot more expression from people who aren't technical. You can ask a question in natural language about the business — not in SQL — and get an instant answer. That's how I prefer to communicate. It reduces the mental load." — Ian D'Silva, Strategy and Operations Lead

The Impact: An entire org running at a different speed

Hex transformed EliseAI's data team from a request queue into a team that builds and operationalizes workflows and adoption followed. Customer-facing teams — regardless of technical skill — are now armed with live data at any time. EliseAI has already grown their Hex subscription by 3x over the past three months.

With Hex, EliseAI achieved:

• 3x growth in self-serve analytics across the company over three months, now 91% of the company accesses data through Hex
• 14 cross-functional teams use Hex
• 7.3K reports and ad hoc analyses built with Hex
• 97% reduction in QBR prep time — from two days to five minutes for a fully customized, data-rich deck
• 98% reduction in audit prep — from two hours to minutes for a complete settings review with live links
• 80% reduction in development time to get an MVP version of a customer audit
• 100+ pre-built slide variations available on demand — a resource that wouldn't exist without automation

EliseAI has embraced agentic analytics across the organization, and they continue to operationalize workflows for customer-facing roles and leverage their learnings to build similar processes for the healthcare side of their business.

“The shift was significant — we went from having a data request queue to having people request access to Hex.” - Hannah Lee, Analytics Engineer