Decimal AI Raises $4M to Bridge Support and Engineering

Decimal AI, a San Francisco-based startup, raised $4 million in seed funding co-led by Khosla Ventures and Kearny Jackson, with participation from Atlassian Ventures, Weekend Fund, and individual backers including former Stripe COO Claire Johnson, to expand what it calls the Customer Engineering Platform, a new category of software designed to bring engineering-level technical understanding directly into customer support conversations, according to the company's own announcement of the round.

Founders Sanjeet Hajarnis and Kevin Raji Cherian, who previously worked together at Eightfold AI, built Decimal around a genuinely specific, well-observed pain point: support escalations traditionally rely on secondhand descriptions of what a product is doing, rather than direct evidence from the product itself, according to Runtime Wire's reporting on the platform's origins. Hajarnis previously led AI at Eightfold and worked on Facebook's News Feed ranking systems, while Raji Cherian built Databricks' vector search product and led infrastructure at Eightfold, giving the founding team genuine technical depth in exactly the kind of complex systems their product is designed to investigate.

Decimal's platform connects directly to a company's repositories, logs, documentation, configuration, production data, and customer histories. When a support ticket arrives, its AI Support Engineer investigates the issue, drafts an answer, takes approved account or billing actions, and can even prepare a bug fix for engineering review, according to Runtime Wire's reporting. That technical scope distinguishes Decimal from conventional support automation, which typically retrieves answers from a static documentation base or knowledge base rather than investigating a product's actual, real-time behavior.

Decimal AI at a Glance

Decimal's growth claims are backed by genuinely concrete, named customer outcomes rather than vague productivity language. Supply-chain risk management firm Resilinc cut its mean time to resolution from 6.5 days to 2.5 days, a 62% reduction, according to Complete AI Training's reporting on the round. Granola doubled its ticket-handling capacity while resolving roughly 70% of common questions through chat before they ever became formal tickets, and separately uses Decimal to automate investigations and generate pull requests that improve its own documentation.

Khosla Ventures founder Vinod Khosla framed the investment around a cost problem he believes hasn't received enough direct attention. "Support is becoming one of software's biggest hidden costs," Khosla said. "Decimal changes those economics entirely with an AI platform that understands the code and customer context as well as the best support engineers." Kearny Jackson co-founder Sunil Chhaya emphasized the founders' specific, earned credibility on this exact problem: "Sanjeet and Kevin didn't build this from the outside. They have the scar tissue and the expertise to close the gap between support and engineering, and that's why some of the most technically sophisticated companies in the market trust them."

Not every account frames this raise as an unqualified validation of the underlying technology at scale. Runtime Wire's analysis was direct about what remains genuinely unproven: "The $4 million round gives Decimal room to turn that improvised process into a product. Its founders now have to prove that access to code and production evidence produces reliable resolutions across customers, rather than impressive demonstrations inside a handful of carefully integrated accounts." That's a genuinely useful reminder that early customer wins, however concrete, don't automatically guarantee the same reliability once the platform scales across a much broader, more varied customer base.

Decimal's raise is worth understanding for any software company evaluating how AI can reduce the costly handoffs between customer support and engineering teams, particularly given the specific, disclosed customer results here, real resolution time reductions and ticket volume handled, rather than generic productivity claims common elsewhere in enterprise AI.

For businesses in technical SaaS specifically, Kearny Jackson's Sriram Krishnan predicted "customer engineering" will emerge as its own distinct technology category, separate from traditional helpdesk software or internal developer tools, a framing worth watching as more companies potentially enter this same space.

What does Decimal AI's platform actually do?
Decimal AI's Customer Engineering Platform gives technical support teams an AI Support Engineer that investigates code, logs, production data, and documentation to resolve customer issues end to end, including answering questions, taking approved support actions, and drafting bug fixes for engineering review.

How much funding has Decimal AI raised?
Decimal AI raised $4 million in seed funding co-led by Khosla Ventures and Kearny Jackson, with participation from Atlassian Ventures, Weekend Fund, and individual investors including former Stripe COO Claire Johnson.

What real results have Decimal AI's customers reported?
Supply-chain risk firm Resilinc reported cutting its mean time to resolution by 62%, from 6.5 days to 2.5 days, while customer Granola doubled its ticket-handling capacity and now resolves roughly 70% of common questions through chat before they become formal tickets.

Decimal AI raised $4 million in seed funding co-led by Khosla Ventures and Kearny Jackson to expand its Customer Engineering Platform, an AI Support Engineer that investigates a company's actual code, logs, and production data to resolve technical support issues rather than relying on static documentation. The company has disclosed concrete customer results, including a 62% reduction in resolution time for supply-chain firm Resilinc and doubled ticket-handling capacity for customer Granola, with overall support interaction volume growing 15-fold since early 2026. Founders Sanjeet Hajarnis and Kevin Raji Cherian, both Eightfold AI veterans, are betting the platform can prove reliable resolution accuracy at broader scale beyond its current handful of deeply integrated early customers.