Summation opens its AI analyst to self-serve teams at $60 per user

Why it matters

Summation's self-serve launch tests whether trusted AI analysis can become ordinary SaaS. Success depends on encoding messy business context without recreating a consulting project.

Summation CEO and co-founder Ian Wong opened the Bellevue, Washington software maker's AI analyst to self-serve customers on September 10th, 2026, letting teams sign up for Summation on plans starting at $60 per user per month.

The availability announcement turns Wong's enterprise analytics thesis into a more demanding product test. Summation spent roughly a year deploying inside businesses including Fanatics, Lineage and Grid, according to the company. Teams can now connect their own data and assign recurring analytical work without first buying a custom enterprise engagement.

Wong founded Summation with CTO Ramachandran "RC" Ramarathinam after the pair worked together at Opendoor. Ramarathinam led Opendoor's core transaction platform. Wong co-founded the online homebuyer in 2014 and served as its CTO through its public-market debut, after earlier working as Square's first data scientist. He holds bachelor's and master's degrees in electrical engineering and a master's in statistics from Stanford.

That background explains the job Wong has chosen for AI. Summation is aimed at the operating reviews, forecasts and follow-up investigations that determine how executives allocate money, inventory and staff. Those assignments require a system to understand a business's definitions and constraints, retain that context between reporting cycles and show where every number came from.

"Most businesses aren't short on valuable problems to solve. They're short on the analytical capacity to solve them," Wong said in the announcement.

From enterprise deployment to self-serve

Summation's Pro plan costs $60 per user per month, while its Max plan costs $200 per user per month. Enterprise pricing is custom, with dedicated deployment support, custom integrations, and advanced governance and security controls.

Summation says customers can schedule reviews, forecasts and reports, then have finished work delivered through email or Slack. Summation says its analyst can monitor changes across products, customers, locations and sales channels, investigate significant movements and trace its conclusions to the underlying data. Summation also says customers can connect through more than 1,200 integrations.

According to Summation, a team can encode data sources, analytical logic and checks into a workflow that runs each week or month. That persistence separates Wong's pitch from the familiar chat-with-a-dashboard interface, where a user still needs to know which question to ask next.

Wong developed the idea from executive meetings at Opendoor. In an essay, he described leadership teams asking why a metric had changed or what would happen if the business shifted spending. Analysts often needed days or weeks to return with an answer, by which point the decision had moved on.

His target is the recurring preparation behind Monday management meetings: pulling data, reconciling definitions, checking calculations and rebuilding the same deck. Summation leaves judgment with employees while assigning that production work to software.

Self-serve exposes the difficult part

The launch broadens Summation's distribution while exposing the central constraint in AI analytics: company data rarely arrives ready for an agent. Metrics carry internal definitions, permissions differ by employee and source systems frequently disagree. A report can be fluent and polished while its calculations rest on the wrong interpretation of revenue, churn or inventory.

Summation describes its answer as a governed context layer containing the customer's definitions, business logic and access rules, combined with model routing and automated verification. Summation says results carry an audit trail back to the underlying data.

That architecture matches a wider shift in enterprise analytics. ThoughtSpot, for example, launched industry-specific versions of its Spotter agent in March 2026 with semantic models intended to ground answers in business and industry rules. Gartner warned in May that agents without semantic and structural context were more likely to produce inaccurate results and waste spending.

Self-serve customers will show how much of that context Summation can establish through software. Large enterprise deployments can lean on forward-deployed engineers and operators to map data, define metrics and identify suitable workflows. A team paying for a self-serve plan will expect a shorter route from connecting a warehouse or spreadsheet to receiving analysis it can use in a meeting.

Summation is therefore selling ease of setup alongside analytical capacity. If customers still need prolonged data cleanup and metric design, the product will retain the economics and pace of an enterprise implementation. If the setup works with limited assistance, Wong can reach finance, marketing, merchandising, revenue and operations teams that would never approve a large deployment.

A $35 million bet on trusted answers

Summation emerged from stealth in October 2025 with $35 million in funding. Benchmark led Summation's seed round, while Kleiner Perkins led its Series A. Kleiner Perkins partner Josh Coyne wrote that Summation's agents had run roughly 7,000 queries and 3,000 model calls in one customer deployment.

Summation says Fanatics used the product to identify more than $10 million in growth and savings opportunities and reduce some reporting work from weeks to hours. Those figures come from a case study published by Summation, rather than an independent performance test, but they illustrate the scale Wong is pursuing: repeated investigations across more combinations of products, markets and operating variables than an analyst group can manually cover.

The self-serve release shifts that bet from a small number of managed deployments toward repeatable software. Wong has spent his career building data systems inside companies where small changes in pricing, conversion and operating assumptions carry large financial consequences. Summation now has to make that discipline usable by teams that were not in the room when its engineers defined the data.