About
The lab behind the patch
SoundSynth AI Studio Inc. operates from Great Northern Way in Vancouver — a neighbourhood where film sound stages meet game studios and hardware hackers. We opened our doors in 2024 with a single conviction: synthetic speech should be authored, not assembled from leftovers.
Origin
Why we built a synthesis lab
Founders Mira Okonkwo and Elias Renard spent a decade in post-production and conversational AI respectively. They kept encountering the same problem — clients paid for "custom" voices that were thinly disguised stock models. The pair leased Suite 106, installed a hybrid rack of GPU clusters and analogue-modelled outboard, and invited engineers who think in patch cables as readily as tensors.
Today the lab employs fourteen people across synthesis research, ear calibration, project routing, and client success. We serve Canadian enterprises, Pacific Northwest startups, and select international brands that value sonic craft over speed-to-default.
Philosophy
Listen first, render second
Algorithms propose; humans dispose. Every voice passes through structured listening sessions where engineers mark micro-timing, plosive weight, and emotional drift. We publish internal QC rubrics adapted from broadcast dialogue editing — because a voice that scores well on MOS tests can still feel wrong in your onboarding flow.
We also believe in transparent deployment. Clients receive documentation explaining how their voice was built, which data was used in training boundaries, and how to disclose synthetic speech to end users where regulations require it.
Studio
Great Northern Way facility
Our suite houses three isolation booths, a main patch floor with 12-channel monitoring, and a client lounge for remote review sessions. Vancouver's creative ecosystem keeps our ears honest — when neighbouring teams push immersive audio standards, we adapt our delivery specs to match.
Visits are welcome by appointment Monday through Friday, 09:00–18:00 Pacific Time. We are incorporated in British Columbia under BN 634081927 RC0001.
Team composition
Research engineers explore timbre architectures and latency budgets. Ear-calibration specialists run structured listening protocols borrowed from broadcast dialogue editing. Project routers keep milestones visible in shared dashboards so procurement and creative leads see the same status without chasing email threads.
We invest in continuing education — AES conventions, Interspeech proceedings, and internal seminars where staff demo patch techniques. That culture matters because synthetic speech moves fast; yesterday's acceptable sibilance profile may sound harsh on tomorrow's smart-speaker firmware.
Community and ethics
SoundSynth AI sponsors student showcases at local polytechnic media programs and contributes pro-bono calibration hours to nonprofit accessibility projects where synthetic narration expands reach. We publish deployment guidance encouraging transparent labelling when audiences interact with generated speech.
Our founders chose incorporation in British Columbia deliberately: provincial privacy rules under PIPA complement federal PIPEDA, giving clients a clear framework when personal data accompanies voice projects. We extend that rigour to our own operations — consent checkboxes on forms are never pre-selected, and cookie preferences require an explicit tap before optional analytics load.
When you walk through Suite 106, you will see patch cables colour-coded like our brand palette — lime for dry signal, cyan for effects sends. That visual language mirrors how we talk about projects: every connection intentional, every route documented.