Some founders fall in love with a product. John Derham appears to have fallen into a long argument with messy information. The setting keeps changing: a bank, a television feed, an enterprise data stack. The irritation remains familiar. Useful facts exist somewhere. They arrive late, stripped of context, trapped in a format no neighboring system understands, or guarded by a pipeline whose creator has sensibly stopped answering messages.
Derham has spent more than three decades close to that irritation. He began in financial services, working in marketing, risk analytics, and decision sciences at Advanta, Royal Bank of Scotland, and JPMorgan Chase. These were places where data was not decorative. A model had to help somebody approve, price, target, or decline something. Numbers were expected to leave the spreadsheet and survive contact with a decision.
That distinction explains much of what came later. Derham's path is less a tour from finance to media to infrastructure than a sustained study of the passage between raw information and operational judgment. He simply keeps finding new places where the passage is blocked.
Act IThe television set becomes a database
By 2009, Derham had left the banking track and founded Probus Media Group as its chief data scientist. Two years later came iQ Media. One account placed the company's beginnings in a Doylestown garage, a pleasingly conventional birthplace for a decidedly unconventional task: make television searchable, measurable, and fast enough to matter while a campaign was still alive.
Television measurement then leaned heavily on scheduled advertising and broad audience estimates. Derham saw another category of value hiding in plain sight. A brand might appear on a player's shirt, behind a news anchor, on the boards at a hockey game, or in a scene nobody purchased as an advertisement. Paid exposure and earned exposure mingled on screen. Most systems were poor at telling them apart, and worse at doing so quickly.
“When I saw marketers struggling with a lack of data around linear TV, I started iQ Media,” he said in 2017. The company recorded broadcast television, indexed it, and built tools to locate brand mentions and images across national and local markets. Its machine-vision work could search for thousands of logos. The intriguing question was not merely whether a logo appeared. It was whether a passive viewer might actually register it. Size, position, and persistence all mattered.
There is something wonderfully literal about the enterprise. Television is a stream. Derham wanted to turn that stream into data, and then turn the data into a decision before the moment cooled. The company stored millions of hours of programming and served customers in sports, automotive, finance, technology, and entertainment. Its public client lists included organizations such as Google, Mercedes-Benz, the NFL, and Sony.
The growth was local enough for Derham to call it “a great Philly story.” In 2015, iQ Media announced a $9 million Series B from Edison Partners and reported a staff of more than 80. Derham was also coaching men's lacrosse at Villanova, where he had once captained the team. The combination suggests a calendar with few ornamental gaps.
In 2018, 4C acquired a stake in iQ Media, and the business was combined with Teletrax to create Kinetiq, a global television intelligence network. Derham moved into chief data and product roles. The technology had traveled from a garage idea to an operation spanning paid, earned, and owned television across local and international markets.
Act IIThe problem beneath the product
A company that records television around the clock learns a few things about data movement. Volume is merciless. Formats disagree. Context gets lost between capture and delivery. A useful answer needs to appear while somebody can still act on it. Beneath every clever measurement sits a less photogenic layer of fetching, transforming, reconciling, monitoring, and moving.
After Kinetiq, Derham founded phoeniQs tech, a managed services and analytics company, in 2020. EASL followed in 2021. Its proposition reaches below a specific application. The company works on the movement layer itself: bring data from a source, change it into the needed form, deliver it to a destination, and maintain a record of what happened along the way.
cloud + legacy + streams
workflow
apps + models + teams
This is a narrow-sounding problem with a talent for expanding. Add a new database and a connector must be configured. Change a schema and a downstream process breaks. Move a workload between cloud and on-premises systems and the exceptions breed. Then AI arrives with an appetite for frequent, high-volume, trustworthy data. The pipeline that was once an internal nuisance becomes the ceiling on the company's ambitions.
Derham's answer is a framework he calls DataDevOps. The name borrows from the transformation software engineering went through when building, testing, deploying, and observing code became a shared and increasingly automated discipline. Data teams, he argues, deserve similar repeatability. Workflows should be tested. Changes should be visible. Governance should be embedded. Recovery should not depend on the memory of the one engineer who understands why a column is named “final_v7_really.”
A business outcome runs across sources, transformations, checks, and destinations. Managing only the connectors misses the story.
Infrastructure should expect schemas, volumes, systems, and requirements to move rather than treating every change as an exception.
Teams need to see failures, lineage, performance, and recovery while the data is still operationally relevant.
Controls and audit trails should travel with the workflow, making safe speed possible instead of turning governance into a queue.
Act IIIAn impatience with fake innovation
Derham's recent writing broadens the argument beyond pipelines. He draws a sharp line between efficiency and innovation. Companies have become adept at optimizing familiar work, trimming cost, outsourcing functions, and layering new interfaces over old rules. These moves can be valuable. They can also preserve the assumptions that made the organization inflexible in the first place.
His phrase for this is “fake innovation.” It is the kind that photographs nicely at launch and sends an invoice to operations six months later. A modern dashboard may still rely on a thicket of fragile transfers. A new AI assistant may sit above data that nobody entirely trusts. The visible layer changes. The penalties for change remain.
This is where his career acquires a pleasing circularity. Banking taught him that information must support a decision. Television taught him that speed and context determine whether information has value. Founding and scaling a media technology company exposed the difficulty of moving enormous, unruly streams. EASL is an attempt to make that difficulty the product rather than the backstage crisis.
- Decision sciences in finance
Advanta, Royal Bank of Scotland, and JPMorgan Chase. - Media becomes measurable data
Probus, iQ Media, and the creation of Kinetiq. - The infrastructure turn
phoeniQs tech, followed by the founding of EASL. - A public case for DataDevOps
Essays on adaptable infrastructure, governance, and the cost of optimization without reinvention.
The operator's lessonBuild where everyone else stops looking
Derham's useful provocation is not that every company should buy another platform. It is that leaders should look below the interface. How many people are manually repairing data flows? How long does a new source take to reach production? Can anyone reconstruct what happened to a record? Does a change create a controlled deployment or an office-wide suspense drama?
Those questions expose the real architecture. They also reveal where talented engineers spend their days. A company can hire more of them, but staffing a brittle system is not the same as improving it. Derham's recent work repeatedly returns to this point: adaptability is a property of the foundation, not a heroic quality to demand from the people standing on it.
EASL's public mission includes the line “empower others and do no harm,” along with commitments to integrity, transparency, and speed. In the abstract, such words can float away like helium balloons at a corporate retreat. Here they fit the product problem. Transparent pipelines expose errors. Governed movement reduces unpleasant surprises. Adaptable workflows return engineers to building instead of babysitting.
The ambition is modestly phrased and technically difficult: make getting data right easier. Derham has been approaching that sentence from different directions since his first analytics roles. He has modeled decisions, indexed television, recognized logos, built teams, raised capital, combined companies, and returned to the pipes.
The pipes, naturally, are still misbehaving. They always will. Systems change because businesses change, and businesses change because reality has the poor manners to ignore a roadmap. Derham's wager is that infrastructure should be built with that rudeness in mind. After 30 years, he is no longer trying to make data sit still. He is trying to make change survivable.