NYTech
People
Personfounder

Olivier Pomel

Olivier Pomel is the co-founder and chief executive of Datadog, a New York-based observability platform he has led since its founding in 2010 [1][2][3][4]. Before starting Datadog he spent nearly a decade in engineering roles, including as vice president of technology at Wireless Generation, senior software engineer at Silicongo, and earlier positions at Neomeo and IBM Research, and he holds a degree from CentraleSupélec [5][6][7][8][9]. His time at IBM Research brought him to the United States in the late 1990s on an internship, initially intended as a six-month stay, after which he remained and later moved into building Datadog in New York [13].

Pomel describes Datadog's origin as a response to the divide between software developers and operations teams, who worked in separate tools and often blamed one another when systems failed; the company's founding premise was to bring both groups into a shared platform with a common view of reality rather than simply building better monitoring tools [11][13]. He has said this same principle of shared language and shared workflows over pure tooling applies to newer organizational splits, such as those between AI engineers and other technical and non-technical staff [11]. He also credits Datadog's seat-agnostic pricing model with removing the incentive customers might otherwise have to limit how many people use the product, since the goal has always been to bring more teams together rather than restrict access [11].

On observability itself, Pomel frames the category as a convergence of what were once separate disciplines including infrastructure monitoring, application performance monitoring, log management, and network monitoring, now unified into a single end-to-end view of how software and business systems behave [10]. He argues that the distinction often drawn between "clear" traditional observability failures and "fuzzier" AI failure modes is overstated, contending that most production issues, AI-related or not, are already ambiguous and hard to diagnose [11]. He identifies three layers where he sees AI creating opportunity: applications built with AI that consume more infrastructure and data, applications built on top of AI models that behave non-deterministically, and the use of AI to automate detection and resolution of operational issues for engineers [10]. He has also described an internally built transformer-based time series model, calling it state-of-the-art not just for observability data but across time series tasks generally, attributing this to the scale and quality signals in Datadog's data, which he says includes ingesting billions of records per second [12].

Reflecting on competitive dynamics, Pomel has said Datadog started in New York at a disadvantage relative to better-funded Bay Area rivals with more experienced founding teams, and argues the company compensated by staying closely tied to customers rather than to prevailing industry narratives [13]. He draws a parallel between early fears that Amazon would dominate the cloud era entirely and later fears that a handful of foundation model companies would capture all value in AI, arguing that in both cases the market proved more distributed than expected once credible alternatives emerged [12]. His advice to founders navigating thick technology stacks is to position close to the customer problem rather than to sit in the middle of the stack, where outcomes are harder to predict [12]. He also distinguishes between a smaller set of AI-native customers building core infrastructure, such as GPU management or foundation models, and the broader base of Datadog customers who are still experimenting with AI and have yet to reach production at scale [13].

Founded

Experience

  1. Co-founder, CEO
    DatadogJun 2010 to Present
  2. Vice President, Technology
    Wireless Generation2002 to Dec 2010
  3. Sr Software Engineer
    Silicongo2001 to 2002
  4. Software Engineer
    Neomeo2000 to 2001
  5. Software Engineer
    IBM Research1999 to 2000

Education

Media & appearances

This page shows public professional information only, each fact cited. Is this you? Corrections or removal within 24 hours, no questions asked.