Tony Jaensch and the Art of Getting the Same Answer Twice

On January 17, 2025, Tony Jaensch dropped into a Medium discussion with a correction. Content Grabber, he explained to a user called DataSurge, had been replaced by Sequentum Enterprise a couple of years earlier. There were more features now, including an Agent Control Center for overseeing large web-scraping operations. He also had something newer to recommend: Sequentum Cloud.

It was a small piece of internet housekeeping. No stage, no launch countdown, just a co-founder keeping the product conversation current. He opened with “Hi DataSurge,” then got down to particulars: browser-based software, data-compliance monitoring, an audit trail. For a business built around collecting information from websites, correcting an outdated description of its own software was an appropriately practical appearance.

The reply also contained an early glimpse of what would become a larger part of Sequentum’s pitch. Jaensch pointed out that AI could be trained to create agents from the text files underlying the company’s Cloud agents. The interesting object was already there: an agent that could be represented as a file, examined, managed and used again. The chatbot could help with construction. The customer would still need something dependable to operate.

By June 2026, Jaensch was back running the company as acting CEO. In July, Sequentum Cloud won a CODiE Award. The quiet product correction and the public leadership announcement belonged to the same continuing story: a software business changing its tools while keeping a close interest in whether the work actually gets done.

Two degrees, then the business of delivery

Jaensch’s education put commerce and computing on the same timetable. At the University of South Australia, he earned a Bachelor of Business with a major in marketing and a Bachelor of Applied Science in Computer Studies. His subsequent career included work at Accenture, IBM and Cognizant, with a focus on strategy, delivery and business development.

Those are useful coordinates for understanding his place at Sequentum. Selling software requires an account of what a customer wants. Delivering software requires an account of what the system can do, repeatedly, under ordinary working conditions. Marketing and computer studies give those two conversations a shared address. His public career brings them together in a business whose output often ends up inside somebody else’s decision.

Sequentum’s history starts in 2008 with Søren Mondrup and Visual Web Ripper. Jaensch joined full-time in early 2014 as co-founder to manage the business while Mondrup built the second-generation product, Content Grabber. That division of work matters. Jaensch’s story is tied to the commercial and operational development of the company, alongside the technical work of his fellow founder.

Content Grabber used a custom Chrome browser. Later came the enterprise platform and, in October 2024, Sequentum Cloud. The names changed as the product range developed. Underneath them remained a familiar customer problem: useful information exists on the web, but turning it into a recurring, usable feed requires considerably more care than copying a page into a spreadsheet.

A difficult website is a business problem

An earlier customer episode makes that work concrete. Retail Insight, a web-based market-research provider, encountered a website that its existing tools and other providers could not handle. Sequentum’s team examined the requirements and built an extraction agent to demonstrate that Content Grabber could do the job. Retail Insight then deployed the software in its own production environment.

The appeal lay in getting past a particular obstruction. A website had frustrated the available tools; a working agent gave the customer a route through. The episode is a useful counterweight to the grand language that tends to gather around data businesses. Sometimes the entire proposition comes down to one awkward site and a customer who needs the information by the next run.

In that phase of Sequentum’s development, the product combined a visual editor with scripting and management tools. Managed Data Services added the option of having Sequentum develop, host, monitor and maintain the extraction agents. Customers could choose how much of the operation to keep themselves. Each arrangement put a different boundary around responsibility for the same recurring task.

For Jaensch, the commercial subject has therefore lasted through several generations of software. A buyer needs access, but also a way to keep that access useful. The glamorous moment is the first successful extraction. The harder business is the succession of ordinary mornings when the customer expects the feed to arrive again.

June 11: back in the chair

On June 11, 2026, Sequentum announced that its board had unanimously appointed Jaensch acting CEO following Sarah McKenna’s resignation. He had been president and co-founder, and had served as the original CEO. The appointment returned an existing company leader to the chief executive role, rather than introducing someone to the business from outside.

It also came with other changes. BC Wilson moved from head of customer success to chief operating officer. Casey O’Neill became chief technology officer after leading AI and engineering. Dave Reinlieb became vice president of information security. Sequentum credited McKenna’s more than eight years of leadership with growth to nearly 100 employees, the launch of Cloud and the integration of AI into the platform.

“I’m looking forward to working with the team,” Jaensch said. His announcement singled out Agent Builder, launched in May, and Checkout Agent, launched in April. The timing put him at the head of a company expanding in two directions: helping people collect web data and helping software agents carry out purchases. Both required the browser to do useful work on somebody else’s website.

There is a practical distinction between returning to a founding idea and returning to a founding job. Jaensch inherited the intervening work of colleagues, an expanded team and a newer product range. The leadership appointments placed customer operations, engineering and security beside the CEO role. The next chapter would have several people’s names on the responsibility chart.

“I’m looking forward to working with the team.”

Tony Jaensch · June 2026

The review comes before the repeat

Agent Builder begins with a plain-language request. A user describes the data required, or supplies a URL and instructions. AI helps produce an extraction agent; the user can inspect and refine it in the Visual Editor. Production execution then uses deterministic logic, without AI inference in the extraction step. The distinction is central to the business Jaensch now leads.

Deterministic execution does not freeze the internet. Prices change, pages move and websites revise their structure. It means the approved extraction logic can be examined and run again without asking a language model to make a fresh judgment on every pass. When something changes, the team has a defined piece of software to investigate. There is a difference between a changing price and a changing method of reading it.

Sequentum records the agent version and run history, along with workflow and network activity. Optional screenshots and video recordings can add context. These are prosaic objects with considerable value when somebody asks why yesterday’s feed differs from today’s. A demonstration wins attention quickly. A log file earns its keep much later, usually when the room has become less cheerful.

In September 2026, COO BC Wilson developed that argument in a paper about using language models in production data pipelines. His concern was the gap between building a quick prototype and operating a dependable recurring process. It is an argument that fits the company’s longer history: make the initial work easier, then give people control over the machinery they depend on.

A prize, a purchase and the next ordinary morning

On July 15, 2026, Sequentum Cloud won the CODiE Award for Best No-Code/Low-Code Platform in the Developer & AI Technology track. The company announced the result two days later. “This is a proud moment for the entire Sequentum team,” Jaensch said. He connected the award with Sequentum’s eighteenth year and the eighth anniversary of its India center.

The award recognized a company effort. Agent Builder had launched before his return to the CEO role, and Cloud before that. Jaensch’s response put the team first, a fitting acknowledgment of the work already in motion. In India, director and general manager Sangeeta Mishra also described the founders joining colleagues for the anniversary celebration. The company’s geography had become part of its working history.

Checkout Agent takes the browser-automation foundation into transactions. Announced in April 2026 with Skyfire integration, it combines Sequentum’s execution tools with identity and payments infrastructure. The proposed uses include buying licensed datasets and renewing digital services. Spending authority and an audit trail become part of the task. Once software can purchase something, a convincing demonstration has acquired a more expensive sequel.

By October, the business was still making its case through everyday customer work. A published retail example described agents monitoring eight competitor websites across more than 50 product categories. Sequentum reported 98 percent extraction accuracy and more than 3,000 hours of annual time savings for that customer. These are case-specific reported results, but they locate the ambition in a recognizable workplace: fewer hours repairing the feed, more time using it.

Jaensch’s public LinkedIn profile places him in the Greater Melbourne Area. Sequentum’s work extends across the United States, Australia and India. Between the founder answering a product question and the executive overseeing that international business sits a persistent obligation: keep the information useful after the excitement of the first run has passed.

His current opportunity is to make the new tools serve that old obligation. AI can shorten the journey from request to working agent. Human review can define what the agent should do. The repeated run is where the promise gets tested. For Tony Jaensch, the next ordinary morning remains a very consequential appointment.