Following Volodymyr Panchenko’s earlier conversation with Grit Daily on empowering SMBs, Portal’s latest launch reframes a fundamental question in artificial intelligence: what comes after chatbots?
The Missing Layer in AI
The AI industry has scaled rapidly, but unevenly. Companies like NVIDIA dominate compute. OpenAI and Anthropic build increasingly powerful models. Yet one critical layer remains underdeveloped: a consumer-facing system that works for ordinary users.
Not developers. Not researchers. Not early adopters. Normal people. Portal One+ is designed to fill that gap.
Rather than offering another chatbot interface, the platform gives each user a private, persistent AI agent that remembers their entire history, accesses real tools such as browsing, file management, code execution, and voice, and develops a compounding understanding of how they think over time.
Unlike conventional systems that reset with every session, Portal’s agents accumulate context.
Day after day. Conversation after conversation.
Technology investor Neeraj Gunsagar, Managing Partner at Dancing Lotus Ventures, described the distinction after spending hours inside the system:
“The real difference is profound. Gemini can pull up your calendar and summarize your emails. That is a search layer on top of your data. Portal builds a compounding understanding of how you think. There is a very profound difference.”
Built for Use, Not Complexity
The founding principle behind Portal is deceptively simple: AI should work without requiring users to understand it. Panchenko built the first version with a specific user in mind, his mother.
“She doesn’t want to know what model she’s running. She wants to say what she needs and have it done. Everything I’ve built in twenty years of tech somehow came down to that.”
This philosophy underpins the product’s design. Users do not manage prompts or models. They interact naturally, while their agent handles execution in the background.
The result is less interaction with AI as a tool and more reliance on it as a system.
Traction That Signals a Behavioral Shift
Early usage data suggests that persistence is not just a feature. It changes behavior.
Portal reports more than 3,200 active users across over 50 countries, with more than 25,000 persistent agents deployed. Its DAU to MAU ratio stands at approximately 42 percent, more than double the typical engagement levels seen in consumer AI applications.
Growth has been entirely organic. The platform reached its first 2,000 users in under 30 days with zero customer acquisition cost. More than 20 percent of users convert to paid subscriptions, driven entirely by word of mouth.
Crucially, the business is already unit economically positive, with revenue per user exceeding compute costs, a rare milestone in an industry where infrastructure expenses remain a primary constraint.
Portal is also in the process of raising a $40 million round, with approximately $11 million secured to date, alongside inbound interest from major AI infrastructure players.
What Users Actually Do
The strongest validation of the system lies not in metrics, but in outcomes.
Portal agents are not just generating content. They are producing economic results.
In one case, a creator used her agent to build a complete monetization funnel, coding a website, setting up payments, writing copy, and executing a launch, despite having no prior technical experience. The result was $15,000 in revenue within 36 hours, eventually scaling to over $49,000.
In another, a user sent a single voice note requesting a song in the style of Fred again… The agent wrote lyrics, composed the track, cloned the user’s voice, and delivered a fully produced song later published on streaming platforms.
A recently laid-off HR professional used her agent to write and publish two books, create seven digital products, and launch a personal website entirely through conversation.
Even edge cases highlight the system’s capabilities. In Madrid, a user avoided a phishing scam when her agent detected a subtle domain discrepancy in an email. In London, another user received a real-time warning about an unsafe pregnancy supplement before her doctor identified the issue.
These examples point to a broader shift. AI is moving from generating outputs to executing workflows.
The Infrastructure Behind Persistence
Portal’s approach is built on a proprietary architecture known as OpenClaw, developed by co-founder and CTO Hamudi Naanaa.
The system is not a wrapper on existing APIs. It is a ground-up infrastructure designed to support persistent agents at scale.
Seven US patents underpin the architecture, alongside more than 100 proprietary algorithms. The platform currently operates across dozens of server shards, coordinating tens of thousands of agents simultaneously.
Despite early exposure to large-scale DDoS attacks and prompt injection attempts, the system has maintained continuous uptime.
But the deeper moat may lie in how learning propagates across the platform.
When one agent acquires a new capability, filing a trademark, publishing music, integrating payments, that capability is tested, refined, and distributed across the network. The first execution may take hours. Subsequent ones can take minutes.
This creates a compounding effect. Every user makes every other user’s agent more capable.
Panchenko describes this dynamic directly:
“The first patent took our agent ten hours to file. The next three trademarks? Ten minutes. That’s not automation. That’s learning. And it compounds across every user on the platform.”
Social Impact and Accessibility
Beyond commercial use, Portal is also experimenting with access.
Through its Portal Next initiative, the company offers free 30-day access to individuals undergoing career transitions, whether due to layoffs, pivots, or new ventures.
The goal is to reduce barriers to entry for people looking to reskill, build, or launch independently. The HR professional who published two books is one example of this cohort.
A New Layer in the AI Stack
The broader strategic positioning is clear. The AI value chain is consolidating around three layers: compute, models, and distribution. While the first two are increasingly dominated by large incumbents, the distribution layer, how AI reaches and serves end users, remains open. Portal is positioning itself as that layer.
“We don’t compete with ChatGPT or Gemini. They’re search engines for the internet. We’re a memory system for your life,” Panchenko said.
The company’s roadmap extends beyond individual users. It is building toward a creator economy for AI capabilities, where users who develop new tools can earn ongoing royalties through a revenue-sharing model.
At the same time, enterprise applications are emerging, including the coordination of large-scale influencer networks through distributed AI agents.
The Founder Dynamic
At the center of Portal’s execution is a tightly defined division of roles between its co-founders.
“My co-founder Hamudi Naanaa built every line of Portal from scratch,” Panchenko said. “I do the product and the people. He does the architecture and the impossible. We don’t overlap. That’s why it works.”
A Shift Already Underway
Portal One+ arrives at a moment when artificial intelligence is shifting from novelty to infrastructure. Its core thesis is straightforward but consequential. Intelligence alone is not enough. Without memory, continuity, and execution, even the most advanced systems remain limited.
“I built Portal so my mom could use AI. Not as a novelty, but as her actual second brain,” Panchenko said. “When she called me during a one-hour outage because she couldn’t function without her agent, I knew we had something real.”
Whether Portal ultimately defines this emerging category remains uncertain. But the direction of travel is becoming clearer. The future of AI may not be defined by how well it responds in a single moment but by how effectively it works over time. And in that future, persistence is not an upgrade. It’s the system.
