I am Sorry

A routine cleaning assignment becomes an unnerving descent into a mansion that never feels quite the same twice. I am Sorry is a live-action, image-driven thriller built around the uneasy idea that familiar rooms can hide new details, new rules, and something deeply wrong beneath their polished surfaces. You play a janitor returning for repeated work sessions, carrying out ordinary tasks while subtle disturbances begin to turn the job into a mystery. The mood is deliberately claustrophobic and observant rather than action-heavy, leaning on dread, discomfort, and the fear of overlooking a vital clue.

Most of the experience revolves around examining rooms, interacting with objects, cleaning assigned areas, and noticing what has changed since the previous visit. A misplaced item, altered decoration, unusual message, or newly accessible detail can push the narrative forward, so careful memory matters as much as conventional puzzle-solving. Story-focused puzzles gate deeper revelations and ask players to connect environmental evidence with the mansion’s growing sense of menace. Progress comes from patient investigation rather than combat or fast reflexes, making it a better fit for a focused solo session with headphones than for players seeking constant scares or arcade-like challenges.

Its use of live-action visuals gives the mansion a disconcertingly tangible quality, especially when repeated spaces begin behaving differently from how you remember them. The central change-detection structure recalls observation horror games such as The Exit 8, but it frames that tension within a more traditional mystery about uncovering the truth behind a location and its inhabitants. Not every optional detail needs to be solved to see the ending, yet digging into the puzzles provides a fuller understanding of what is happening. Fans of slow-burn psychological horror, FMV-style presentation, and escape-room mysteries should find plenty to engage with, while newcomers should expect a deliberately measured pace that rewards attention over brute-force experimentation.