Backup is an unfortunate word. It arrives wearing a beige cardigan, asks whether anyone has tested the recovery plan, and clears the room before the coffee cools. Joe Gaska hears something livelier. He hears a business remembering itself. A price changed. A customer left. A field was added to a database. An order moved. Each edit looks small at the moment it happens. Together, the edits form a movie, and Gaska has spent much of his career insisting that companies should keep the film.
His shorthand is crisp: measure change over time. The phrase fits a founder who studied applied mathematics and computer science at the University of Maine at Farmington, then built two enterprise-software companies around streams of events. It also fits the Cape Cod native who describes himself as having salt water in his veins. Tides are change with a timetable. A shoreline is a versioned document.
Gaska’s present title is co-founder and chief product officer of GRAX. It is a different seat from the CEO role he held while building the company. In late 2025, GRAX combined with CapStorm under AscendX, bringing together two businesses that help enterprises control Salesforce data outside Salesforce itself. Ted Pappas took the chief executive role. Gaska moved closer to product, where his long-running argument now has to become software at a larger scale.
The first company was already about a trail
Before GRAX, there was Ionia Corporation. Gaska founded it in 2004 and led it for a decade. Ionia became a Salesforce and Heroku Platinum Partner, delivering work to more than 300 customers in fields including manufacturing and life sciences. Its connected-product business helped companies turn machines into services by collecting what those machines did and making that activity useful.
LogMeIn acquired Ionia in 2014. The company’s work became a foundation for Xively, LogMeIn’s Internet of Things platform. The fashionable term was IoT, but beneath it sat a plain problem: a connected object creates a history, and somebody has to capture, interpret and act on that history. A machine without its trail is merely blinking. A machine with a trail can explain itself.
After the acquisition, Gaska worked with Xively before returning to founder mode. By early 2017 he was thinking about a different kind of connected history: the records locked inside cloud applications. A Salesforce account might hold years of sales, service and customer activity. The application made those records easy to create, but extracting every version, preserving its context and keeping it ready for another use was awkward.
His complaint was philosophical before it was technical. A customer could pour data into a subscription service and then find itself effectively renting access to that data. Gaska wanted the customer to keep an independent, usable copy in its own cloud environment. The distinction sounds procedural until the service changes, an employee deletes a record, an auditor asks what happened, or a data team wants to train a model on yesterday rather than today.
The past gets a second job
GRAX launched in 2018 with backup and recovery as the practical doorway. Gaska’s larger thesis was that stored history should do more than wait for catastrophe. A conventional backup is an insurance policy in a locked drawer. A usable history can reveal how pipeline velocity changed, when a customer relationship began to weaken, or which version of a schema produced a misleading answer.
He likes the word trajectory. A snapshot can tell an executive where the business stands. A sequence tells the executive which way it is moving. This is why the rawness of the record matters. If old data is flattened, stripped of context or stored somewhere difficult to reach, the most interesting questions become expensive before anyone can ask them.
Artificial intelligence made the second job more obvious. A model learns from examples, and a company’s cloud applications contain years of cause and effect: the quote that became an order, the case that became a renewal, the change that preceded a problem. Gaska was talking about that training set before generative AI turned every earnings call into a thesaurus exercise. His warning remained practical. A model cannot learn from corporate memory that was never kept, cannot be reached, or sits under another vendor’s control.
Market pull, then money
For its early stretch, GRAX was bootstrapped. Gaska and his co-founder, Christopher Haley, put money into the company, reinvested revenue and did not pay themselves salaries. By 2019, Gaska wrote that demand was pulling faster than the business could reinvest. Volition Capital supplied a $12.8 million Series A.
The announcement contained the standard ambitions and one useful joke. GRAX would not spend the money on a lavish office with a Starbucks barista in the lobby. Every penny, he said, would go into scaling the team, product and go-to-market work. “Then we can get the barista!” It was a wink, but also an allocation memo. Raise because the market has exposed a constraint. Spend against that constraint. Foam art can remain on the roadmap.
His public manner mixes technical density with homely comparisons. In one live discussion, Gaska compared interest in backup to interest in toilet paper: almost nobody gets excited until the shelf is empty. It is an engineer’s sales instinct - translate an invisible risk into the household moment when everyone finally notices.
The engineer never quite leaves the room
Gaska is an executive who can still descend into the machinery. His public credentials include Salesforce administration, consulting and development certifications. His writing shifts easily from ownership principles to tutorials about moving Salesforce data into Tableau, Excel and Google Sheets. The grand argument is always attached to an ordinary user trying to get a report out before lunch.
That range helped put him on the Dreamforce main stage and brought Ionia and GRAX recognition inside the Salesforce ecosystem, including a Salesforce Innovation Award. It also explains why his language tends to alternate between architecture and agency. Architecture decides where the records live, how often they move and whether the chain of custody survives. Agency decides whether the customer can use those records without asking a platform owner for permission.
The combination is a useful founder habit. Gaska does not argue for data ownership as a decorative principle. He ties it to the next action: restore the missing record, reconstruct the old schema, compare one month with another, or hand a clean historical set to a model. A principle earns its keep when it changes what a user can do on Tuesday afternoon.
There is a certain applied-mathematics neatness to the whole career. A business at one moment is only a point. Keep enough points and you have a line. Keep the line and you can see direction, speed and interruption. The product opportunity is hidden in the difference between the point and the path.
The hour when nobody thought about business
There is another Gaska anecdote, and it contains no databases. During remote work in 2021, a GRAX employee suggested a virtual visit to a goat sanctuary. Colleagues were skeptical. Gaska loved it. The staff watched goats wander for nearly an hour while the sanctuary owner explained the farm and delivered goat trivia.
At the end, she asked everyone to baa. People looked mortified and reached for mute. She called them back. Slowly the employees, Gaska included, made goat noises together. Everyone cracked up. He later said he regretted not booking more time because, for a minute, nobody thought about business. He called it probably the best thing he had planned as CEO.
Leadership lesson, delivered on four hooves: a brief shared absurdity can restore more attention than another serious meeting about attention.
The story is revealing because Gaska’s normal vocabulary is so earnest: chain of custody, retention, schema, governance. Yet an organization is also a history of odd little moments that people choose to remember. The executive obsessed with preserving digital memory understood the morale value of making one ridiculous human memory on purpose.
A title changes; the thesis stays
The 2025 combination with CapStorm placed GRAX inside a broader enterprise-data business. CapStorm brought near-real-time replication and self-hosted Salesforce integration. GRAX brought its record of historical versions, backup, recovery and reuse. The common premise was customer control across a company’s own cloud or on-premises environment.
Gaska’s move into the product role is therefore less a departure than a return to first principles. The chief executive has to carry the entire company. The product chief can keep asking the question that started the company: what should a customer be able to do with its own history?
His career offers a modest lesson for builders. Preserve optionality before its use is obvious. Keep the original record. Keep the context around it. Let customers hold the keys. Significance is a late-arriving guest, and it is embarrassing when you have thrown away its chair.