Datadog, Inc. is an American technology company that provides an observability platform for cloud-scale applications, offering monitoring, security, and analytics for servers, databases, tools, and services through a software-as-a-service (SaaS) data analytics platform. Founded and headquartered in New York City, the company went public on the Nasdaq exchange in 2019 and has since become a major player in cloud monitoring. As AI adoption grew, Datadog expanded its platform to support monitoring of AI workloads, including model performance and data pipelines, positioning itself as a key observability provider for machine learning operations.
Datadog focuses on helping engineering teams manage the reliability, performance, and cost of their infrastructure and applications. It offers dashboards, alerting, and visualizations of metrics, with a broad ecosystem of integrations that supports major cloud providers, container orchestrators, and AI development tools. As of 2025, the company supports over 750 integrations, including for Artificial intelligence and Machine learning frameworks.
History
Datadog was founded in New York City in 2010 by Olivier Pomel and Alexis Lê-Quôc, who met as undergraduates at Ecole Centrale Paris and previously worked together at Wireless Generation, an educational technology company. After Wireless Generation was acquired by NewsCorp, they sought to create a product that reduced friction between developers and systems administrators, who often worked at cross-purposes. They built Datadog as a cloud infrastructure service with dashboards, alerting, and metric visualizations, quickly filling a need as cloud adoption increased.
In 2015, Datadog opened a research and development office in Paris, expanding its engineering footprint. A year later, it moved its New York headquarters to the New York Times Building to support a team that doubled over the year. Full-stack monitoring arrived with the beta of Application Performance Monitoring in 2016, which was followed by growth into other areas.
The company went public with an initial public offering (IPO) on September 19, 2019, selling 24 million shares and raising $648 million, valuing it at $8.7 billion. Prior to the IPO, Datadog had rejected a $7 billion acquisition offer from Cisco. The IPO saw the stock rise about 37% on the first day, giving the company a market value near $10 billion. Datadog was added to the S&P 500 Index in July 2025, reflecting its maturity and financial standing.
Acquisitions
Datadog has grown through a series of strategic acquisitions to broaden its platform capabilities. In April 2021, it acquired Sqreen, a cloud application security company, bringing runtime application security and protection features to the platform. This added an AppSec layer that could detect threats in real-time, useful for AI applications where input manipulation can be a risk.
The company acquired Codiga in April 2023, and static analysis platform, expanding into developer lifecycle tools and code quality. This supports monitoring of AI codebases and automated scrutiny of code for edge cases, which is helpful for maintaining model invocation quality. In 2025, it expanded further with two acquisitions: Metaplane (April 2025), a data observability company, and Eppo (May 2025), a feature flagging and experimentation platform.
Through Metaplane, Datadog enhanced monitoring for data pipelines and AI workloads, addressing the need for data quality checks to prevent model drift. Eppo integrated feature flagging and experimentation capabilities into the product analytics library, useful for testing AI models in production environments. These acquisitions reflect a strategic path to supporting the full AI development lifecycle.
Products and Services
Datadog offers a range of observability product lines, each designed to support cloud and hybrid environments. Core products include Infrastructure Monitoring, Network Performance Monitoring, and Network Device Monitoring, which provide real-time visibility into health and performance. Serverless Monitoring extends this to function-as-a-service workloads, crucial for AI inference services that often run on serverless architectures.
Cloud Cost Management helps businesses track spending across cloud resources, including compute, storage, and data transfer, which are significant in training large neural networks. The platform generates dashboards, sets alerts, and visualizes metrics, enabling teams to respond quickly to anomalies. It integrates with major AWS, an and Google Cloud, as well as other providers.
For AI-specific needs, Datadog introduced support for model monitoring. This is through the platform, though details are sparingly apparent. The company acquired Metaplane to handle data lineage and quality, while Eppo brings experimentation, which is useful for A/B testing models with different hyperparameters. As Generative AI spreads, Datadog helps monitor generative models for accuracy, responsiveness, and adherence to service levels.
Technology
The Datadog agent, installed on hosts and containers, collects and sends metrics to the backend. The agent is written in Go, based on a rewrite since version 6.0.0 released on February 28, 2018. It originally used Python (for which it hooks used) but moved to Go for better performance. The backend leverages state of the art software like Apache Cassandra, PostgreSQL, and Kafka, plus D3 for data visualization.
For machine learning, Datadog's backend supports integrating with popular ML frameworks, though the company does not offer model training. Instead, it functions as an observability layer for AI systems. For example, network performance to monitor inference clusters, and that agent can be extended for custom metrics. The platform uses proprietary data analytics to detect anomalies.
Models that are deployed on Apple devices or Arm based chips can be monitored via integrations that capture system metrics. Datadog does not train AI models; instead, it positions itself as a tool to ensure that AI applications built with large language models (LLMs) or other approaches operate reliably, helping with issues such as response latency and token consumption.
Funding and Financials
The funding history shows strong investor confidence. In 2010, Datadog launched with a seed round including participants like NYC Seed, Contour Venture Partners, IA Ventures, and Jerry Neumann and Alex Payne. A $6.2 million Series A followed in 2012, co-led by Index Ventures and RTP Ventures. Series B came in 2014 led by OpenView Venture, and in 2015 Series C of 31 million led by Index. In 2016, ICONIQ led a $94.5 million Series D, one of the largest funding allocations for a New York company that year.
The company's IPO set an official valuation of $8.7 billion, and the stock surge pushed the market cap to near $10 billion by the close of trading. Follow-on offerings and stock price growth have defined its public tenure. As of 2024, Datadog had approximately 6,500 employees in 33 countries, with offices in New York, Boston, Denver, San Francisco, Paris, Dublin, Amsterdam, Sydney, Tokyo, and Singapore, and about 59% of employees based in the U.S.
AI Monitoring and the Ecosystem
Datadog's expansion into AI observability comes as part of a broader trend in monitoring platforms. While companies like AWS Trainium and Groq provide hardware accelerated inference, Datadog offers the software telemetry to track performance and usage. For AI workload, it provides dashboards that combine infrastructure metrics with model-specific metrics, such as token usage and confidence scores, helping teams diagnose bottlenecks and maintain [[machine-learning|machine][learning]] accuracy.
The integrations with OpenAI and Anthropic are likely, but even without such, Datadog captures API calls and their costs. In context of Generative AI, monitoring can include content moderation to ensure outputs against policy. Datadog's ability to handle streaming data, for token-by-token, uses this method in a backend. Its platform helps mitigate model deployment issues, like model- drift.
Related companies in the AI observability space include startups such as Essential AI, but Datadog distinguishes itself by its infrastructure data. It also aligns with cloud providers like Microsoft Azure, Amazon Web Services, and Google Cloud to help their AI services. While Datadog does not provide the models itself, it allows Machine learning teams to ensure that applications built on Large language model are reliable.
Impact and Future Directions
Datadog has uncertainty become a standard tool for cloud monitoring, its AI focus reflects the industry's shift. In 2025, it added the Eppo and Metaplane acquisitions, indicating a move to become an asset in the AI development lifecycle. With AI applications being deployed in mission-critical, observability is crucial.
It is likely that Datadog will deepen its AI capabilities, and the company is, synthetic-ai trends, but an official place for its strategy is not fully known as of early 2025. It continues to expand integrations each month, embracing support for Artificial intelligence frameworks and possibly adding features like monitoring for LLM-specific metrics. The future may see more model-observability products.
In summary, Datadog's revenue has grown as it has added, and its transition to AI-ready monitoring is both a strategic opportunity and a difficult region. With competitors like AWS Trainium and Groq not at this observability area, Datadog remains the default choice for developers. Its S&P 500 inclusion in 2025 reflects its financial stability, and its AI focus are likely to drive innovation in the realms of Machine learning deployment, model reliability, and data quality.