Decimal AI Raises $4M to Build Customer Engineering

A customer opens a technical support ticket with one sentence. The answer may be hiding across source code, production logs, feature flags, account data, configuration, documentation, and a decision an engineer made 6 releases ago. Decimal AI has raised $4M to make that investigation a support capability instead of another interruption waiting for engineering.

The September 15, 2026 Seed round was co-led by Khosla Ventures and Kearny Jackson, with Atlassian Ventures and Weekend Fund participating. Decimal is also introducing what it calls a Customer Engineering Platform, built around an AI Support Engineer that works across customer questions, technical evidence, support actions, and bug fixes prepared for engineering review.

The financing is small beside the nine-figure rounds collecting attention across AI. Its operating claim is larger: support teams can own more customer resolution when they receive the evidence behind product behavior, not merely a polished summary of the ticket.

What Decimal AI raised

Decimal's official announcement identifies the financing as a $4M Seed co-led by Khosla Ventures and Kearny Jackson. Atlassian Ventures and Weekend Fund joined the round, while individual backers named in the release include Claire Johnson, Michelle Valentin, and Rimple Patel. Axios separately corroborated the amount and institutional investor group.

Decimal did not disclose a valuation, financing instrument, ownership terms, investor check sizes, board changes, or total capital raised. The company says it has used the new financing to expand the range of technical issues its system can investigate and to deepen integrations across the support and engineering tools customers already use.

That wording matters because Decimal is describing capital already moving through the product rather than a distant hiring or expansion promise. Its careers page currently lists 3 San Mateo roles across engineering and go-to-market, but the release does not provide a headcount target, spending allocation, or revenue plan.

Why technical support keeps reaching engineering

Traditional support automation is good at retrieving known answers. Technical B2B software creates a harder class of ticket: the customer describes a symptom, the documentation describes expected behavior, and the actual answer depends on what happened in that customer's environment at a specific moment.

Resolving that problem can require querying logs, inspecting code paths, checking a configuration, matching an account state to a deployment, and reviewing earlier conversations. The work exists between customer service and engineering, which is why support teams often escalate precisely the issues that matter most to the customer and cost the most engineering attention.

Decimal's thesis is that an AI system can perform much of that investigation inside the support workflow. Its AI Support Engineer can analyze code, logs, configurations, production data, documentation, and customer history; draft an evidence-backed response; take defined support actions; and prepare bug fixes for engineering review. The company calls the resulting operating model customer engineering.

What the early customer evidence shows

Decimal says it launched in March 2025 and now works with Granola, Resilinc, Tealium, BuildOps, Lucidworks, and an unnamed Fortune 5 technology company. It reports that the volume of support interactions resolved through the platform grew 15-fold from the beginning of 2026 through the funding announcement.

The most useful detail sits inside the customer workflows. In Decimal's Granola case study, the system investigates tickets across CloudWatch logs, code, documentation, and customer context before the support team begins its response. Decimal reports that Granola doubled ticket capacity, resolves 70% of common questions in chat before they become tickets, and uses generated pull requests to repair documentation gaps.

In the company's Resilinc case study, mean time to resolution fell from 6.5 days to 2.5 days while the system helped validate issues before escalation. These are company-reported customer results, not independently audited operating data. They still show the product's intended economic lever: give support teams more technical context before engineering becomes the default path to an answer.

The founders are building from the handoff

Decimal was founded by Sanjeet Hajarnis, co-founder and CEO, and Kevin Raji Cherian, co-founder and CTO. The company says the pair met at Eightfold AI, where Hajarnis led AI and Raji Cherian led infrastructure for large-scale matching systems. Its announcement also credits Hajarnis with earlier work on Facebook's News Feed ranking and Uber pricing systems, and Raji Cherian with building Databricks' vector search product.

Those backgrounds are relevant because Decimal's product has to reason across several layers at once: application behavior, customer context, production evidence, access controls, and the language a support team can safely send back to a user. Its security materials describe read-only access by default, evidence citations, tenant isolation, configurable retention, and Zero Data Retention agreements with AI providers. Enterprise adoption will depend on whether those controls stay credible as the system takes on more investigations and support actions.

What the investor group is underwriting

Khosla Ventures and Kearny Jackson are backing a new category claim as much as a product. Decimal argues that customer engineering can become a distinct function in the same way go-to-market engineering emerged around increasingly technical revenue work. The bet is that complex software will keep shipping faster while customer questions become more specific, more contextual, and less answerable from a static knowledge base.

Atlassian Ventures adds another useful piece of the map. Jira and related collaboration tools often sit inside the handoff among support, product, and engineering. Its participation does not establish a commercial partnership, but it places a strategic investor close to the workflow Decimal wants to reorganize.

The Seed round gives Decimal more room to widen the issues its system can resolve and the systems it can connect. The consequential evidence will arrive inside live support organizations: which investigations can move without an engineer, which actions customers trust, and whether the facts scattered across a software company can reach the person responsible for the answer before the escalation queue does.