Todd Persen, CTO of Hydrolix, explains how the company handles massive real-time log data for observability and CDN and edge use cases, why he joined after previously building observability and log companies (including an exit to ServiceNow), and what Hydrolix is focused on next. He also shares how the team runs globally as a remote org, why shorter release cycles matter, how product and engineering alignment depends on clear goals and visible prioritization, and where AI tools like Claude Code fit today without removing humans from the loop.
Todd Persen on Hydrolix, global scale log data, and running engineering in hypergrowth
In this episode of Software Leaders Uncensored, host Steve Taplin speaks with Todd Persen, the CTO of Hydrolix. Steve introduces Hydrolix as a company with over $145 million raised to date and frames it as a fast growing player in the enterprise data space. Todd describes Hydrolix as a platform for real-time data at global scale, with a focus on log management and adjacent observability use cases where data volumes are massive, including CDN and edge scenarios.
What Hydrolix does and where it fits
Todd explains that Hydrolix centers on handling large scale, real-time log data, particularly in environments with extremely high volume. He mentions customers and partners dealing with CDN and edge challenges and says the company is positioned for observability style insights where logs are produced at massive scale.
Todd’s path to Hydrolix
Todd says he has worked in this space for a long time. He references his earlier company InfluxData, where he helped build the time series database InfluxDB and worked in metrics before the term “observability” was widely used. He describes InfluxDB as open source and mentions building open source momentum and navigating monetization through an open core model.
Todd then describes founding a company called Era, focused on logs in a space similar to Hydrolix, and says Era was acquired by ServiceNow in 2022. After deciding to leave ServiceNow, he joined Hydrolix, explaining that he was already familiar with the company due to overlap in the broader space. When asked about Hydrolix CEO Marty Kagan, Todd says that founders in the same market tend to know each other and that he had built a relationship over time, even though their prior companies were not going after the same end user.
Todd also shares that joining Hydrolix offered a different kind of stage experience for him. He describes his earlier ventures as more “zero to one” and says Hydrolix was past that phase, which made it interesting to contribute during a rapid growth stage.
The biggest technical surprise after joining
When Steve asks about surprises, Todd says the biggest technical surprise was that the technology and platform were already really good. He contrasts this with the expectation that many database startups have “skeletons in the closet.” He describes Hydrolix’s approach as leveraging object storage to achieve low cost and high performance through engineering design strategies. He also says the platform was already working at scale when he arrived.
Todd references a public example: Hydrolix handled CDN logs for the Super Bowl last year through Fox, describing ingest rates around a petabyte and a half per day, which he equates to roughly a terabyte per minute.
What the team is working on
Todd says Hydrolix is expanding beyond its established position in the CDN market. He notes a significant partnership with Akamai and says Hydrolix works with Akamai customers to provide observability and insights tied to CDN products. From there, he describes exploring additional markets and use cases, including potential security use cases and other edge scenarios outside CDN.
He explains that expansion is not always about building a brand new core platform, but about identifying the right persona and adjusting what already exists so the new use cases are reachable for go-to-market teams. He mentions cross-functional work with sales, marketing, and product to support that shift.
Todd also says customers regularly ask about AI strategy. He describes interest in using AI related approaches for things like improved anomaly detection, more proactive insights and alerts, and improving data onboarding, especially as data sources and log formats multiply. He notes that teams can no longer assume a single person understands all log types entering the system and says tools that can extract patterns and shapes from data in a more automated way are useful, including where LLMs can help.
Engineering team structure and operating globally
Todd says Hydrolix is a remote company with a few small offices, including a small office in Portland, where CEO Marty Kagan is based. He describes the workforce as globally distributed, influenced in part by the long-term Akamai relationship and the need to support customers across regions. He says core engineering is largely in the US and Europe, and he mentions an engineering team in Ukraine.
On coordinating across time zones, Todd says Hydrolix tries to avoid too many calls requiring everyone to attend. He mentions a weekly tech leader sync scheduled early in Pacific time to better accommodate Europe and the US. He also notes flexibility in start and end times when needed, including occasional late calls to connect with teams in APJ, but says the goal is to build a team that does not require heavy meeting overhead. He states that distributed teams need trust and that remote work is a bad fit if teams require babysitting.
What slows teams down in hypergrowth
Todd describes Hydrolix as a company of a little over 200 people, with roughly 40 percent in core engineering and additional technical staff in customer success engineering and SRE functions. He estimates technical staff overall may be 50 to 60 percent of the company.
He says one of the biggest slowdowns in hypergrowth is that organizational structure takes time to catch up, which can create overlap between teams or missing coordination layers. Todd says Hydrolix started as an engineering-first company and got far without traditional product leadership, with some people doing double duty across engineering and product. He says the company is now seeing the need for more traditional product structure and that building stronger product and engineering alignment is a 2026 goal.
Todd references his background at Pivotal Labs, describing it as a strong agile environment with pairing, test-driven development, and a peer relationship between product and engineering.
Product and engineering alignment, priorities, and release cycles
Todd says product reports up through a separate leader and views that separation as healthy. He emphasizes that product cannot define an effective strategy without engineering input, and engineering cannot wait on product without engaging on constraints and possibilities. He says alignment starts with clear organizational goals and that prioritization issues often require leadership involvement rather than leaving trade-offs solely to individual engineers or product managers.
He also discusses the value of making priorities visible through tools and boards and mentions Pivotal Tracker features such as velocity-based estimates and release marker changes that make trade-offs obvious.
Todd shares a process shift toward shorter release cycles, citing benefits like faster customer value, faster feedback, and stronger engineer ownership. He says there is no substitute for code being used in production, and that long release cycles can reduce clarity because people forget the details of what they shipped and why.
Tech debt and AI workflows
Todd says tech debt is always present, even in very small startups. He adds that debt can be healthy when taken on intentionally to achieve another goal, but becomes a problem when it is forgotten. He says tech debt has not been one of Hydrolix’s biggest issues, though it exists like it does for every team.
On AI in engineering, Todd says Hydrolix has provided Claude Code licenses to the engineering team and describes adoption as still early, with some engineers using it more than others. He says the company still relies on automated testing and manual release testing because the system is complex and distributed. He adds that Claude Code has been useful for certain work like small bug fixes and UI projects, but he would be surprised if AI generated code is even in the double digits as a share of total code.
Communicating with non-technical stakeholders and closing advice
Todd says communication works best when it accounts for the stakeholder’s background. He mentions using analogies, specific data points, and clear explanations of trade-offs. He says complicated issues can be broken down into understandable parts and that patience plus clarity usually brings people along.
For his closing advice, Todd says tech leaders should build organizations made up of people they trust. He ties trust to remote work success and recommends being transparent, being honest, trusting the team, and not giving up.
