Side, a seasoned services provider formerly known as PTW, has entered into a six‑month memorandum of understanding (MOU) with Danish artificial‑intelligence specialist Modl.ai to co‑develop an AI‑driven quality‑assurance (QA) platform for video games. The agreement binds the two companies to collaborate on a system that blends Side’s extensive QA experience with Modl.ai’s cutting‑edge AI technology, aiming to create a solution that augments, rather than replaces, human testers.
**Background of the partners** Side has long been a fixture in the outsourcing and testing arena, offering a broad suite of services to game developers worldwide. Modl.ai, by contrast, is a relatively new entrant to the gaming market. Founded in 2018, the Copenhagen‑based startup secured $1.7 million in seed financing in 2020 and has since focused on applying machine‑learning techniques to automate repetitive testing tasks. Both firms see the partnership as a chance to combine their respective strengths: Side’s deep domain knowledge and client relationships with Modl.ai’s innovative AI models.
**The vision behind the AI‑enabled QA system** Martin McBride, Side’s chief commercial officer, describes the joint effort as “the best of Side and the best of Modl.ai in a single, unique offering.” He adds that the gaming industry has been searching for a "holy grail"—a reliable way to use AI to enhance QA services without sacrificing accuracy. According to McBride, previous attempts at AI‑based testing have fallen short, but the new approach promises to finally bridge that gap.
Historically, AI‑driven QA tools required developers to hand over entire game builds or even source code so the AI could analyse the software from the inside out. Modl.ai’s earlier solution relied on an integration plugin and SDK, which forced developers to expose internal assets and created considerable friction. Founder Christoffer Holmgard acknowledges that this model “became difficult over time” and limited adoption. The revamped system eliminates the need for deep integration.
Instead, developers provide the game exactly as they would to a human tester—typically a compiled build that runs on a target device. The AI then interacts with the game by watching the screen, simulating inputs, and recording every observable event. A second AI layer reviews the generated data, flags potential defects, and escalates items that require human judgement. **Building “game knowledge”** The first phase of the workflow is what Modl.ai calls "building game knowledge." Developers upload assets, design documents, screenshots, and other relevant materials to a cloud‑based portal.
Using natural‑language processing, the AI can converse with the team about gameplay mechanics, level layouts, and expected behaviours. Holmgard explains that the interaction is entirely conversational: "We just chat with the system about our game, but we also feed it images, design docs, and other context." Once the AI has a baseline understanding, the development team defines test objectives. These can range from traditional test cases—such as verifying menu navigation or collision detection—to more exploratory scenarios that leverage the AI’s pattern‑recognition capabilities. Holmgard stresses that human expertise remains essential at this stage: "You need to craft prompts and language that guide the AI to do the right things." **From AI‑generated runs to human‑validated bug reports** When the AI executes a test run (referred to as a "session" by Modl.ai), it captures a wealth of data: textual logs, video snippets that serve as visual evidence, and performance metrics.
This information is compiled into a bug report that is handed over to a senior QA specialist. The specialist reviews the evidence, decides whether the issue is a genuine defect or an intended feature, and determines the appropriate remediation path.
Distinguishing bugs from intentional design choices is one of the trickier aspects for an autonomous system. Holmgard notes that ongoing feedback from human testers helps the AI refine its judgment.
Over time, the model learns to associate certain visual cues or gameplay patterns with either bugs or features, improving its precision. **Career evolution and the “human‑in‑the‑loop” ethos** Side sees the partnership as an opportunity to upskill its workforce. McBride envisions QA staff transitioning into "AI engineers," responsible for configuring prompts, interpreting AI outputs, and integrating findings into development pipelines.
This shift, he argues, can elevate salaries, create new career trajectories, and even bring jobs back to regions that have lost testing work due to cost pressures. According to McBride, AI will handle the most repetitive, low‑value tasks—such as smoke testing each new build—freeing testers to focus on strategic analysis, creative test design, and higher‑level problem solving.
He describes the result as a "more fulfilling career" because employees can engage in work that adds greater insight rather than merely ticking boxes. Benedikte Mikkelsen, Modl.ai’s chief product officer, echoes this sentiment, stating that many current tests are monotonous and could be automated, allowing humans to concentrate on assessing gameplay feel and conducting inventive testing approaches. Both leaders stress that AI is not a replacement for human QA but a tool that moves people up the value chain. McBride points out that testing costs remain low in some regions, making a full AI takeover economically impractical.
Instead, the hybrid model leverages AI to reduce error rates and workload while preserving human oversight. **Scope and limitations of the technology** Side’s chief operating officer Oliver Deslandes clarifies that the solution is presently best suited for simpler titles—puzzle games, card games, and other experiences with repeatable mechanics. Fast‑paced 3D open‑world or fighting games present challenges that the current iteration cannot yet handle.
Deslandes emphasizes that early adopters should be developers willing to experiment and iterate alongside the technology, acknowledging that the journey will involve pilots, refinements, and gradual expansion to more complex genres. **Looking ahead** The six‑month MOU serves as a testbed for refining the AI‑QA workflow, gathering real‑world data, and proving the value proposition to potential clients. If successful, Side and Modl.ai hope to broaden the platform’s capabilities, eventually supporting a wider array of game types and scaling to larger development studios. In summary, the partnership blends Side’s seasoned QA operations with Modl.ai’s AI expertise to create a system that automates routine testing, enriches human testers’ roles, and maintains a steadfast "human‑in‑the‑loop" philosophy.
By doing so, the two companies aim to set a new standard for quality assurance in the gaming industry, offering developers faster feedback cycles, reduced costs, and higher‑quality releases while preserving the essential judgment that only experienced humans can provide.