Side, a seasoned services provider formerly known as PTW, has entered into a six‑month memorandum of understanding with Danish artificial‑intelligence specialist Modl.ai. The agreement sets the stage for a joint effort to build a next‑generation, AI‑driven quality‑assurance (QA) platform aimed at the gaming industry.
While Side brings decades of experience in delivering QA services to publishers and developers, Modl.ai contributes cutting‑edge machine‑learning technology that it has been refining since its 2018 launch and its $1.7 million seed round two years later. Martin McBride, Side’s chief commercial officer, describes the partnership as a marriage of complementary strengths. “We’re combining Side’s deep QA expertise with Modl.ai’s AI capabilities to create a solution that could become a ‘holy grail’ for the games sector,” he says.
“No one has yet cracked a truly effective AI‑augmented QA workflow, and we believe we’re on the cusp of doing just that.” Historically, AI‑enhanced QA tools required developers to hand over full game builds—or even source code—to the testing vendor. Modl.ai’s earlier offering relied on an integration model that used a plugin and SDK, demanding that developers share internal assets. This approach introduced significant friction, as founder Christoffer Holmgard admits, and over time proved cumbersome for both parties. The new system eliminates the need for any internal access.
Holmgard explains that the process mirrors how a human tester would receive a game: the product is delivered on a device that closely matches—or is identical to—the end‑user’s hardware. The AI then runs the game, watching the screen, simulating inputs, and logging every observable event. A secondary AI layer reviews this data, flags potential defects, and escalates them to a human reviewer for final validation.
The first phase of the workflow is what Modl.ai calls “building game knowledge.” Developers upload assets, design documents, screenshots, and other relevant materials, then converse with the AI in natural language to teach it the fundamentals of their title. Holmgard notes that this dialogue can include text, images, and even video, allowing the system to develop a nuanced understanding of the game’s mechanics, narrative, and visual style. Once the AI has a solid grasp of the project, developers specify the testing objectives.
These can range from classic regression test cases to novel scenarios that are specifically tailored for AI interpretation. “Human expertise is still essential here,” Holmgard says.
“You need to craft prompts and guide the AI so it performs the right actions and focuses on the right areas.” During a test run—referred to by Modl.ai as a “session”—the AI captures a wealth of evidence: textual logs, video snippets, and raw telemetry. This package is handed off to a seasoned QA specialist who reviews the findings, decides which issues merit attention, and determines the appropriate remediation steps. Holmgard emphasizes that distinguishing a genuine bug from an intended feature can be challenging for an autonomous system, and human feedback is critical for refining the AI’s judgment over time.
Feedback loops are built into the platform. When testers label a finding as a bug, a non‑bug, or provide context for why a behavior is intentional, the AI ingests this information and gradually fine‑tunes its detection algorithms. Over months of iterative training, the system becomes more adept at recognizing subtle design nuances and reducing false positives. Side envisions a broader transformation for its workforce.
Rather than remaining confined to repetitive manual testing, QA staff can evolve into AI engineers who manage, train, and interpret the outputs of the machine‑learning models. McBride believes this shift will elevate the value chain, offering higher‑paid, more intellectually rewarding roles and potentially repatriating jobs to regions where low‑cost testing has previously migrated. “By automating the drudgery of smoke tests and routine verifications, we free our people to think strategically, analyze results, and contribute to higher‑level quality decisions,” McBride explains.
“That leads to more fulfilling careers and better outcomes for our clients.” Benedikte Mikkelsen, Modl.ai’s chief product officer, echoes the sentiment, noting that many current tests are monotonous and could be performed more efficiently by an algorithm. “When the AI handles the repetitive flows, humans can focus on creative testing—evaluating gameplay feel, exploring emergent behavior, and conducting exploratory sessions that add real value,” she says. Both executives stress that AI will not replace QA professionals but will augment them. The price pressure on low‑cost testing hubs makes a complete substitution impractical, yet the technology can raise the overall skill level of the workforce.
McBride describes AI as a “sliding scale” where the human element remains indispensable: “Our ethos is that there will always be a person in the loop. The tool amplifies human capability, reduces error, and lifts the whole process.” Regarding applicability, Side’s chief operating officer Oliver Deslandes cautions that the current solution is best suited for simpler titles—puzzle games, card games, and other experiences with repeatable mechanics. Fast‑paced, graphically intensive 3D open‑world or fighting games remain a longer‑term goal. “We’re targeting clients willing to embark on this journey with us, accepting that the technology will mature as we iterate together,” Deslandes says.
In summary, the Side‑Modl.ai partnership aims to create an AI‑enhanced QA pipeline that respects the need for human oversight while leveraging automation to tackle repetitive tasks, accelerate bug detection, and open new career pathways for QA professionals. The six‑month MOU will serve as a proving ground, with pilots expected to demonstrate the system’s potential across a range of game genres and lay the groundwork for broader adoption in the future.