Side, a seasoned 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 that will integrate Modl.ai’s AI capabilities into Side’s quality‑assurance (QA) processes, creating a hybrid system that blends machine speed with human judgment. ### Background and Strategic Fit Side has built a reputation over many years for delivering outsourced services to the gaming industry, ranging from functional testing to full‑cycle support. Modl.ai, on the other hand, is a relatively new player that entered the games market in 2018 and secured $1.7 million in seed funding two years later.
While Modl.ai’s early AI‑enabled QA tools required developers to hand over entire builds—or even source code—through a plug‑in and SDK, this approach proved cumbersome and created friction for studios that needed rapid iteration. Martin McBride, Side’s chief commercial officer, describes the MOU as “the marriage of Side’s deep testing expertise with Modl.ai’s cutting‑edge AI technology.” He adds that the collaboration aims to deliver what many in the industry call the “holy grail” of QA: an AI‑augmented workflow that can automatically surface bugs while still allowing seasoned testers to make the final call. ### How the New System Works According to Modl.ai founder Christoffer Holmgard, the revamped solution eliminates the need for any internal code access.
Instead, developers provide the game exactly as they would to a human tester—installed on a device that mirrors the end‑user’s hardware. The AI then runs the game, monitors the screen, simulates key inputs, and records every interaction.
This raw data is passed to a second AI layer that analyses the session, flags anomalies, and generates preliminary bug reports. The first phase of the workflow is what Modl.ai calls “building game knowledge.” Developers upload assets, design documents, screenshots, and other relevant materials. Through a natural‑language interface, they converse with the AI, teaching it the rules, mechanics, and objectives of the title. Holmgard explains that this dialogue can include text, images, and even video snippets, allowing the system to develop a nuanced understanding of the game world.
Once the AI has a baseline comprehension, developers specify the testing objectives. These can range from traditional test cases—such as verifying menu navigation or level loading—to more exploratory scenarios that are tailored to the strengths of the AI. Holmgard emphasizes that human expertise is still required at this stage to craft effective prompts and guide the AI toward the desired outcomes. After a test run, the system compiles a comprehensive report that includes: - Detailed textual descriptions of each detected issue; - Short video clips that serve as visual evidence; - Full log files and telemetry data.
These artifacts are handed over to a human QA specialist, who reviews the findings, validates the bugs, and decides on the next steps in the production pipeline. ### The Human‑in‑the‑Loop Philosophy One of the most challenging aspects of automated testing is distinguishing a genuine bug from an intentional feature. Holmgard notes that continuous feedback from human testers helps the AI refine its judgment.
Over time, as testers label reports as “bug,” “not a bug,” or “acceptable design choice,” the AI’s internal models become more attuned to the subtleties of game design. McBride expands on this philosophy, describing AI as a “sliding scale” that never fully replaces human input.
He argues that the technology should elevate QA staff from repetitive, low‑value tasks to roles that require higher‑order thinking. By automating smoke tests, regression checks, and other routine procedures, testers can focus on strategic analysis, creative exploration, and the kind of problem‑solving that adds real value to a project.
### Career Evolution and Economic Impact Side sees the partnership as a pathway for its employees to become “AI engineers.” This shift could re‑skill workers, allowing them to command higher salaries and opening new career ladders within the company. McBride believes that by moving talent up the value chain, Side can also repatriate jobs to regions where they have previously disappeared due to cost pressures.
In his view, the AI‑enhanced workflow makes it feasible to retain skilled staff in higher‑wage economies while still delivering cost‑effective services to clients. Benedikte Mikkelsen, Modl.ai’s chief product officer, reinforces this point, stating that many current tests are monotonous and could be performed more efficiently by machines. She envisions a future where human testers concentrate on assessing gameplay feel, narrative cohesion, and other creative dimensions that machines are not yet equipped to evaluate.
### Scope and Limitations Side’s chief operating officer Oliver Deslandes cautions that the technology is presently 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. Deslandes stresses that early adopters should be willing to pilot the system, accept iterative improvements, and understand that the solution will evolve based on real‑world feedback.
### Looking Ahead Both companies acknowledge that the journey is just beginning. The six‑month MOU provides a testing ground to refine the AI models, expand the range of supported game genres, and demonstrate tangible cost and quality benefits to clients. If successful, the partnership could set a new industry standard for AI‑augmented QA, proving that intelligent automation and human expertise are not mutually exclusive but rather complementary forces that together raise the bar for game quality.