Manny Medina just pulled off one of the most impressive seed rounds of the year. The Outreach founder's new startup Paid closed an oversubscribed $21.6 million seed led by Lightspeed, pushing the London-based company past a $100 million valuation before even reaching Series A. That's a stunning vote of confidence in Medina's bet that AI agents need an entirely new way to charge customers - one based on actual results, not seat licenses.
The numbers tell the story of just how desperate the market is for a solution to AI's billing problem. Paid has now raised $33.3 million total - including a €10 million pre-seed in March - and sources tell us the valuation has already crossed $100 million. That's before the company has even announced a Series A.
The timing isn't coincidental. As AI agents flood the enterprise market, companies are discovering that traditional software pricing models don't work. Per-seat licensing makes no sense when your 'employees' are algorithms. Usage-based pricing could bankrupt agent makers who pay their own fees to model providers and cloud platforms.
'Per-user fees don't work because agent makers pay usage fees to the model providers as well as to cloud providers,' Medina told TechCrunch. 'Unlimited use could drive them into the red.'
The solution Paid is building essentially turns every AI agent into a performance-based contractor. Instead of paying for the privilege of using software, companies pay for specific outcomes - cost savings, revenue generation, efficiency gains. It's a fundamental shift that mirrors how the best human consultants have always charged.
'You need to show the value the agent is delivering to your customers, because agents are running in the background for the most part,' Medina explains. 'If you're a quiet agent, you don't get paid.'
That philosophy is already winning over customers. Artisan, the viral sales automation startup that's been making headlines with its aggressive 'stop hiring humans' marketing, uses Paid's platform. The company just landed ERP vendor IFS as another customer, signaling that established SaaS players see agents as their next growth engine.
The market validation goes deeper than just customer traction. Lightspeed's Alexander Schmitt says his firm has invested 'more than $2.5 billion into AI infrastructure and application layer companies over the last three years' and witnessed the carnage firsthand. A recent MIT study found that 95% of enterprise AI projects deliver no value, with only 5% making it to production.
'The core of that problem is that no one can really attach value to what agents are doing today,' Schmitt said. Companies don't want to pay agents to produce more emails no one reads or generate reports that gather digital dust.
Medina's track record certainly helped close this round. Outreach, his previous company, reached a $4.4 billion valuation by automating sales workflows. But the context has shifted dramatically since then. The SaaS boom was built on predictable recurring revenue from human users. The AI era demands proving value before extracting payment.
Schmitt thinks Paid has cracked something unique here. 'It's something that we haven't seen someone else build,' he said. But if results-based billing really does unlock mass agent adoption, expect plenty of competition to emerge.
The round included participation from new investor FUSE and existing backer EQT Ventures. With $33.3 million in the bank and a proven founder at the helm, Paid is positioning itself as the infrastructure layer that could finally make AI agents economically viable for enterprises.
For Medina, it's a bet that the future of software isn't about what it can do, but what it actually delivers. In a market flooded with AI promises and pilot programs, that might be exactly the reality check the industry needs.
Medina's massive seed round signals that investors believe we're hitting an inflection point where AI agents move from experimental pilots to core business infrastructure. But success will ultimately depend on whether Paid can prove that results-based billing actually accelerates agent adoption beyond the current 5% production rate. If it works, we could be witnessing the birth of an entirely new software pricing paradigm. If not, even $33 million won't be enough to solve AI's value problem.
