Project
Years of AR, MR and VR built into live guided tours — Boston, Lexington, Nashville, Atlantic Canada, Rome, Pompeii, Venice.
Overview
See Reality builds immersive experiences for tour operators, municipalities and historical attractions, and integrates them into tours those clients are already selling. I am one of the developers on them: across the work below I contributed to interaction design, Unity implementation, iteration against what guides reported back from the street, and the unglamorous part of getting a build to survive a tour day. Where this page says our team, that is what it means — the Venice multiplayer racing work is the one piece that is mine, and it is in progress.
None of it is a demo. These run on a schedule, in public, at a ticket price, operated by staff who did not build them, for families, children, older adults and people who have never worn a headset — often twenty at a time, outdoors, with pedestrians walking through the scene. The guide is not being replaced: the XR is one chapter of somebody else’s two-hour experience and has to hand the group back on time.
Across several years that has covered room-scale mixed reality on Quest, mobile AR triggered from street signage, tablet content a guide holds up, cinematic 360° on lightweight AR glasses, hand-tracked physical interaction, historical reconstruction registered to real streets, five-language localisation, and — currently — networked multiplayer in Normcore.
Everything I have worked on or contributed to is below, grouped and named, including the experiences whose final client I cannot confirm.
Deployments
Eight operators, six with a page of their own. Each one is a tour that was already selling tickets before any of this existed.
Five mixed-reality experiences carried on a two-hour guided tour along the Freedom Trail, and the operation my earliest See Reality work fed into. It sells publicly and holds 4.9 out of 5 across 83 TripAdvisor reviews. Everything is registered to the real place: the market lines up with a colonnade that is still standing, and the cannon sits on the plaza’s actual bricks while strangers walk past the muzzle.
Seven experiences built for one of the larger operators working Rome and Pompeii, sold into their tours as a paid immersive add-on: meeting Julius Caesar, fighting gladiators from the floor of the real Colosseum arena, four temples of the Roman Forum, an ancient Pompeii market, and Pompeii before and after Vesuvius. At Roman volumes the failure modes that matter are throughput and device turnaround between groups.
A municipal client rather than a private operator, which changes the brief. Lexington runs a walking-tour programme with more than twenty-five guides, so the target was three experiences any guide could add to a tour they already give, without rehearsal. Visitors stand where buildings no longer exist and see them rebuilt, meet figures including Anna Harrington, and are put inside moments from April 1775. Free mobile AR off QR codes on city signage covers the visitor who turns up with a phone and no ticket.
Four experiences our team designed and built into a single guided walk: the city’s first settlers at Fort Nashborough, its historical architecture, early Music City, and the rock-and-roll era of the clubs the group is walking past. The scenes layer onto the physical city while the guide keeps talking, so the cut points are set by the storytelling rather than by the software.
XR built into two live tours in Atlantic Canada: a historic walking tour, and a food-and-drink tour where the content is tied to what is being eaten and where it came from. Around six experiences across the two itineraries — guests step onto a 19th-century lobster boat and deal with the crew, meet the figures behind Canadian Confederation, walk a market of the Anne of Green Gables era, and run into dinosaurs. The design question here was pacing: every segment has to end cleanly enough for the guide to pick the group back up.
A cinematic 360° experience delivered on lightweight AR glasses and localised into five languages, and an XR gondola simulator a participant paddles through a stylised medieval Venice while the guide carries on managing the group in the real world. I am currently building a multiplayer gondola racing experience off the back of it, networked with Normcore — in development, and nothing about it is deployed.
An XR experience built around Niagara tourism, for a different operator and a site that behaves nothing like a city walking tour.
Fan-experience work from the sports end of the studio’s client list rather than the heritage end.
Relive 1776
The five experiences on the Boston tour. The cannon is the one I point at in interviews, for reasons under it.
It teaches a visitor to work a historical field gun by hand and in order — charge, ball, rammer, aim down the plaza, linstock, report — and it does not fire if a step is missed or taken out of sequence. A real procedure, decomposed into physical steps, wrapped in feedback that lets a stranger finish it unaided. That is the same problem as simulation-based training, minus the assessment, and it is the most transferable thing in this whole roster.
A reconstruction of the market that stood on the ground the group is standing on: stalls, traders and hanging goods laid over Quincy Market’s real colonnade. The interaction vocabulary is one gesture — reach for the thing — because nobody is teaching a control scheme to a first-time user in a public square.
An immersive recreation of the ride, from inside it rather than watching it happen.
An immersive historical experience built around the events running into the American Revolution, staged on the streets where they happened.
Revolutionary-era Boston and the men organising in it, sited in the part of the city where that organising went on.
Additional experiences
Other See Reality work I have contributed to. Four of these have no deployment name I can confirm, so they are described without one rather than given an invented client.
The stones rebuilt at full height. In the studio’s own capture they are standing in the middle of the Faneuil Hall plaza, against a building that is still there — which is the clearest demonstration I have of what scale does once you put it in a street people already know.
A 360° immersive piece built against lightweight AR glasses rather than a headset. Glasses change the brief completely: no hand tracking, no room scale, and a visitor who can still see the guide the entire time.
An immersive reconstruction of the temple, built from historical source material.
An immersive historical experience centred on the founding of the country.
An experience carried internally under that name.
For younger visitors
Children are not a lesser version of this audience. A nine-year-old in a borrowed headset on a public street has no patience for onboarding, no fear of breaking anything, and an adult next to them who wants to see it work inside five seconds.
A dragon that passes through the real trees above the square and breathes fire, at a size that reads from the ground. The ceiling is the actual sky, which is a scale no screen gets to use.
A scaled-down version of the cannon experience for younger visitors.
British regulars, dancing — the children’s counterpart to the Revolutionary-era content the adult tour carries.
A dinosaur encounter built for younger visitors.
What the work involves
The through-line across all of it, and the part that transfers to XR training and simulation.
The population these experiences serve has never used one, so the setup is part of the design rather than overhead around it: a guide kneeling to fit a headset onto a child is the first ten seconds of the experience, performed in public with a group waiting. Nothing may need explaining, nobody should be able to perform an interaction badly in front of strangers, and per-person onboarding time gets multiplied by twenty.
See Reality’s own answer is History Glasses: roughly 400 g, held up to the eyes like binoculars rather than strapped on, so a visitor never loses sight of the guide, the group or the street. One hardware choice carries the entire argument about what XR is for on a tour — it is a chapter of someone else’s experience, not a replacement for it.
Hand tracking here is an operational choice before it is an elegant one. There is no controller to hand over, explain, drop, or lose between groups, and a visitor’s own hand reaching into the scene needs no tutorial. The cost is that tracking has to hold up outdoors, in daylight, over a full day of use.
Location-based XR lives or dies on alignment. The market has to run the length of a colonnade that is still standing; the cannon has to sit on the plaza’s bricks. Get it wrong and the effect does not degrade politely — it stops being the place and becomes a screensaver.
Reconstruction is the other half of the same idea: a visitor stands on the ground a building occupied and then stands inside it. It puts the design work in the source material — what the building actually was — and in making a complicated historical space legible at a glance to someone who has thirty seconds.
The visitors around that gladiator fight are the tour, not a test audience. Everything here has to survive a schedule: staff who did not build it, groups back to back, mixed hardware, no reliable network, devices charged and cleaned and accounted for, and recovery a guide can perform in fifteen seconds in front of a paying group.
The standard I took from this work and now apply everywhere: a build is not finished when it demonstrates well. It is finished when somebody who did not write it can run it twelve times a day without me. That covers performance on standalone hardware, URP pipelines tuned for passthrough, and behaving the same across a fleet that is not on identical firmware.
The Venice racing work is the first networked thing I have built here: several participants in one shared scene, synchronised state, on operator networks, with guests who joined at different moments. It is the same step a training experience takes when it stops being one trainee at a station and becomes a team running a procedure together.
See Reality is a tourism and location-based entertainment studio — not an industrial training company, and nothing here certifies anyone in anything. What transfers is the method: translate a real process into an ordered interactive sequence, teach the action physically, guide a novice with no mental model, give feedback in the moment, and fit the whole thing inside a workflow somebody else owns. I have done that repeatedly, for paying strangers, in several countries.
Next
Case study A wedding recorded from every direction, so nobody is stuck with one person’s edit of it.
A family museum you enter on foot — built from interviews, tested against flat photographs.
Case study A card game an older adult designs the rules for, and their grandchild has to follow.
A running coach that guides by rhythm instead of telling you to speed up.
Palette