Emergent, a fast growth vibe coding startup, announced the launch of Wingman, an autonomous personal agent designed to accomplish low stakes and annoying tasks using everyday tools.
Since its launch last year, Emergent has seen explosive adoption, claiming a user base of eight million builders across 190 countries. Their vibe coding platform allows people to describe an application into existence without touching a line of Python or Javascript. Now, CEO Mukund Jha wants to apply that same frictionless philosophy to the mundane administrative overhead of being a human in the digital age.
“Most people aren’t failing at productivity. They’re buried under the smaller tasks that never stop coming,” said Mukund Jha, co-founder and CEO of Emergent. “We proved with software creation that the right technology, built the right way, reaches everyone. Wingman applies that same principle to autonomous agents. Now, anyone can have an always-on team working in the background, not just people who know how to build one.”
The Trust Boundary Problem
The biggest hurdle for the burgeoning”agentic economy hasn’t been intelligence, it’s been trust. We’ve all seen the horror stories of early autonomous agents hallucinating a polite “yes” to an expensive flight or accidentally CC’ing a CEO on a snarky internal draft.
Emergent is attempting to solve this with what they call “trust boundaries”. Unlike a “set it and forget it” bot, Wingman categorizes tasks by stakes. If it’s clearing out spam or organizing a calendar invite with a close friend, Wingman acts autonomously. But if the agent needs to send a mass email, modify a production database on GitHub, or commit a significant amount of money, it pauses and pings the user for approval via a notification.
It’s a pragmatic middle ground between the totally manual Copilots of yesterday and the sometimes-reckless autonomy of experimental GitHub repos.
Where You Actually Live
Interestingly, Wingman isn’t another tab you have to keep open. It lives inside the messaging apps where most people already spend their time: WhatsApp, Telegram, and iMessage.
By integrating with existing communication stacks, Wingman feels less like a new software tool and more like a very competent, very quiet remote assistant. It connects to many of the world’s most popular productivity apps like Gmail, Outlook, Slack, and Salesforce via standard sign-ins. There’s no developer setup or API key shuffling required, a move clearly aimed at the prosumer market rather than just the tech elite.
The agent also features a persistent memory, meaning it remembers your tone, your preference for afternoon meetings over morning ones, and the fact that you’re allergic to certain airlines. It’s an attempt to solve the “Goldfish Effect” of LLMs, where every session feels like a first date.
The Agentic Arms Race
Emergent is entering a crowded field. With OpenAI, Google, and Anthropic all racing to turn their models into agents, and hardware plays like Rabbit and Humane still licking their wounds from 2024, the software-first approach seems to be the winning bet.
With backing from a who’s who of Silicon Valley, including Khosla Ventures, SoftBank, Lightspeed, YC, and Google’s AI Futures Fund, the company has the capital to fight for the pole position in the agentic era.
If Wingman can do for personal admin what Emergent did for app deployment, we might finally be moving away from the era of “AI as a toy” and into the era of “AI as a utility.”
Anyone can try Wingman today at emergent.sh.
