Side, a veteran services provider formerly known as PTW, has entered into a six‑month memorandum of understanding (MOU) with Denmark‑based artificial‑intelligence specialist Modl.ai. The agreement sets the stage for a joint venture to build an AI‑enhanced quality‑assurance (QA) platform aimed at the video‑game industry. While Side brings decades of experience in outsourced services, Modl.ai contributes a newer, data‑centric approach to game testing, having secured $1.7 million in seed capital after its 2018 launch. Martin McBride, Side’s chief commercial officer, describes the partnership as a marriage of complementary strengths.
“The MOU effectively brings the best of what Side has and the best of what Modl.ai has together in a unique solution,” he says. “One of the crown‑jewel, holy‑grail experiences for the games industry right now has been using AI technology to better augment QA services. No one’s really found a great solution for that.
We think we have.” Historically, AI‑driven QA tools required developers to hand over full builds or even source code, often through cumbersome plugins and SDKs. Modl.ai’s earlier offering, for example, demanded deep integration that created friction and limited scalability. Founder Christoffer Holmgard notes that the new system eliminates that barrier. “We take the product as you would give it to a QA tester, as you would eventually give it to the consumer,” he explains.
“You load it onto a device that is identical or very similar to the end‑user’s hardware. Our solution then watches the game play out on the screen, simulates key actions, records every visual and telemetry event, and passes that data to a second AI layer that flags potential bugs, escalates them for human review, and asks whether they merit further action.” The workflow begins with what Modl.ai calls “building game knowledge.” Developers upload assets—textures, models, design documents, even screenshots—and converse with the AI using natural language.
This dialogue helps the system form a semantic model of the game world, mechanics, and intended player experience. “All of this is in natural language,” Holmgard says.
“We just chat with the system about our game, but we’re also adding images, design docs, and other artefacts.” Once the AI has a baseline understanding, developers specify test objectives. These can range from classic regression cases to novel scenarios that play to the strengths of machine‑learning‑based exploration. Holmgard emphasizes the need for human insight at this stage: “You need some human expertise to adapt the prompting and the language to the AI agent so it does the right things.” The AI then runs a “session” (Modl.ai’s term for a test run), generating bug reports that include textual descriptions, short video clips as evidence, and comprehensive log data.
The reports are handed to seasoned QA specialists for validation. “We get all the texts, and we get some video snippets that we call evidence along with all the log data and everything,” Holmgard notes.
Human reviewers decide whether each flagged item is truly a defect, a design feature, or a false positive, and they provide feedback that the AI uses to refine its future judgments. Over time, this feedback loop sharpens the system’s ability to distinguish subtle design nuances from genuine bugs.
Side envisions a broader shift in the industry’s talent landscape. By automating repetitive, low‑value tasks, QA professionals can transition into “AI engineers” who fine‑tune prompts, curate training data, and interpret high‑level analytics.
McBride argues this upskilling moves staff up the value chain, potentially repatriating jobs that migrated offshore due to cost pressures. “Instead of just doing testing and climbing the ladder from junior tester to senior tester, they can become an AI engineer,” he says. “That allows our people to earn more, opens new career paths, and lets us bring jobs back to regions where they historically disappeared.” The partnership also promises a more fulfilling day‑to‑day experience for testers.
Repetitive smoke tests and routine verifications often bog down workflow, limiting creative problem‑solving. With AI handling those chores, testers can focus on strategic analysis, gameplay evaluation, and inventive testing methodologies.
“One of the things that this tool does very well right now is that it reduces the burden of repetitive tests,” McBride explains. “By knocking these out and automating them, they get to think more strategically and grow their careers.” Benedikte Mikkelsen, Modl.ai’s chief product officer, echoes this sentiment: “There are some very tedious tests that testers are performing right now.
You don’t really need a human doing the same flow over and over again. That person could be doing something much more valuable, like assessing the gameplay itself or doing more creative kinds of testing. That, I think, is a value proposition of using AI.” Both executives stress that AI is not a replacement for human QA but a catalyst for augmentation. McBride points out that testing price points remain low in certain regions, making a full AI takeover impractical.
“There will always be an opportunity for people to do a lot of the testing work because the price point of testing right now is so low in certain countries that it’s very hard to augment it completely away,” he says. “But by giving our people tool sets like this, I’m allowing them to further their career, have a longer career outside, and keep them for longer by moving them up the value chain.” Side’s philosophy treats AI as a “sliding scale” where human oversight is non‑negotiable.
“There will always be a human in the loop—that’s sort of our ethos,” McBride affirms. “This tool really does allow our people to take the best of what AI can offer, reduce the burden, and also reduce mistakes. A lot of this work is tedious and prone to human error.
You take that out of the equation, and everything ratchets up a notch.” When asked about the types of games the solution can currently handle, COO Oliver Deslandes admits the technology is best suited for simpler titles—puzzle games, card games, or other experiences with repeatable mechanics—rather than fast‑paced 3D open‑world adventures or fighting games. “We have to focus on titles that are more appropriate,” he says.
“What we’re aiming for at this stage is clients willing to join us on the journey, because the results and cost‑savings will vary depending on the project’s complexity.” Deslandes concludes that the partnership is a work in progress. “The reality is that it’s a journey, but if you are at the table now with us and maybe doing some pilots, then you will see where this tech goes.” The MOU thus marks the beginning of a collaborative effort to blend human expertise with machine intelligence, hoping to reshape QA workflows, elevate talent, and ultimately deliver higher‑quality gaming experiences.