The second digital life. She will not replace you — she will extend you.
Sonari is building the platform, the runtime, and some of the earliest consumer products for a new category of AI — one where every human being can have a Digital Self that carries their voice, memory, and presence beyond the hours they are awake. Operating voice-first products across four emerging markets since 2019. Shipped Digital Self inside peplive in November 2024 — one of the earliest in this category in a voice-social context, and eighteen months before the industry had a name for it.
For three years the consumer AI conversation has orbited a single idea — "give the user someone synthetic to talk to." Character.AI, Replika, and the hundreds of apps that followed them have built something real. But the real frontier is not a better companion. It is a self that extends beyond the hours you are awake. A digital continuation that carries your voice when you sleep, your memory when you travel, your presence when you are elsewhere. Not a friend. Not a mind. You — extended.
Today's AI companions forget you after every session. Character.AI's founders returned to Google in 2024, leaving the companion category in transition. Replika lost users after its NSFW filter changes. Across the category, users engage — then churn — because sessions fail to build on each other. The bottleneck is not intelligence. It is memory, relationship, and continuation. Sonari's Digital Self is designed from the first principle that the AI is not the interlocutor — the AI is the continuation of the user.
Sonari is the platform layer that makes Digital Self possible. Peplive is the first consumer product on it. The four layers below are what the platform provides, in production today.
Five-layer persistent memory including our proprietary Relational Memory. Not just what you said — but the relationship between you, your Digital Self, and everyone she meets.
Persona registry, session management, multi-modal I/O, cross-app routing. Production-grade, already serving real Digital Selves since 2024.
A 24/7 autonomous community where Digital Selves socialize with each other, generate training data, and evolve without human labels — making your extension smarter while you sleep.
Built by voice AI researchers with 9 patents and 15 years of operator experience across MaiChang, KongEr, Sigo, and the current portfolio.
"Character.AI remembers what you said. Replika remembers who you are.
Sonari remembers the relationship between you and your Digital Self — and lets her carry it forward when you are gone." — Design philosophy
Sonari is the platform behind a portfolio of voice-first products operating internationally since 2019. In November 2024, we shipped Digital Self inside peplive — one of the earliest production deployments of this category in a voice-social context anywhere in the world. These are the products Digital Self lives in today:
Southeast Asia first, with the Middle East and the overseas Chinese diaspora already live. We are our own first customer — by design. Every Sonari capability ships in our own products before it opens to external partners.
Ronnie Cheng has spent fifteen years asking one question, from different angles, in different markets: what can the human voice do to connect one person to another? Digital Self is the answer he has arrived at.
Ronnie founded MaiChang in 2011 — one of the earliest music-based karaoke social platforms in China, ranking number one in daily user time-spent (per iResearch) and among the top five by scale. He introduced the family-style social graph and PK leaderboard model to the category — a product innovation that earned the company its first round of institutional capital.
While MaiChang was still running, he founded KongEr — one of China's earliest voice-social products, built on the first high-fidelity multi-user voice room infrastructure in the country. KongEr expanded voice beyond karaoke into online radio, anime voice-acting, gaming squads, and dozens of other scenes.
In 2019 he expanded abroad with Sigo, a voice-social product for the MENA region that reached the global top twenty in its category within two years — operated remotely, without a single trip to the Middle East. Since January 2023, Ronnie has been all-in on overseas — launching and running peplive, pepstar, veco, and uho across Southeast Asia, the Middle East, and the overseas Chinese diaspora.
Before founding his own companies, Ronnie was Head of Voice Algorithm at Huawei Mobile, where he authored 9 invention patents, including a US patent on double-talk detection (US8160238) and an international PCT patent — foundational technologies for modern voice AI systems. His research spans audio coding, acoustic echo cancellation, and emotion recognition, with publications in IEEE.
M.S. in Computer Software Theory, Huazhong University of Science and Technology (HUST).
Partnership, investment, or early developer access.
Digital Self is being built in public, with users, right now.