From frontend lead at a state enterprise to AI flagship for short-stay hosts. The studio's range is narrow on purpose — software, AI, and the connective tissue between them.
A studio is the sum of what it ships. Six engagements, six clients, six different scopes — every one shipped to launch.
Our first AI-native product. Generated personalised guidebooks for short-stay hosts, then answered guest questions 24/7 — a chatbot trained on the property, not the internet. Built and shipped end-to-end.
Built the whole stack across three on-demand businesses sharing one rider network. Customer + driver + merchant apps, dispatch, payments, ops back-office. From spec to launched. The business has since wound down; the build itself is the catalyst engagement that started Bytemotion.
Frontend lead on the customer mobile apps and the back-office. Discovery, route planning, reservations, and payments — for the state authority running Thailand's EV rollout.
Frontend on the employee platform: room booking, leave requests, internal comms. Multi-tenant, used by partner state enterprises too.
Built VIMS, the back-office module that lets vendors plug into Hubswitch, import their stock, sync two-way with the main system, and run inventory operations from one place.
Bytemotion-owned cross-platform desktop binary (macOS + Windows) that reads Thai ID cards and pipes the data straight into any web admin. Used in production by Puredent dental clinics and embedded inside Hubswitch flows. Free to try.
Multi-channel order operations, fulfilment integrations, and a new direct-to-consumer storefront. Live work under NDA — we'll write it up once the client is ready to share.
Wi-Fi password. How does the AC work. Where's the nearest 7-Eleven. Most hosts answer those questions ten times a week, in three time zones, on no sleep.
We built a structured guidebook the host edits in plain language, then layered an agent that retrieves from it — the property, the neighbourhood, the guest's actual reservation. The agent doesn't guess. It quotes the host.
Web app for the host to compose and publish. Retrieval over the guidebook plus reservation context. A guest-facing chat that opens from the booking link. An eval suite the host can run before publishing — every change tested against fifty stock questions.
The retrieval logic, the eval cadence, the human escape hatch — that's where the work is. The model swaps out every six months. The product is everything around it.