Accel Doubles Down on Fibr AI as Agents Turn Static Websites Into One-to-One Experiences

While advertising and audience targeting have become increasingly personalised, the website — often the final destination for that traffic — has largely remained static. Fibr AI is working to close that gap by using AI agents to transform generic webpages into one-to-one experiences tailored to individual visitors. This vision has led Accel to deepen its investment in the company.

Accel has led Fibr AI’s $5.7 million seed round, following an earlier $1.8 million pre-seed investment in 2024. The latest round also included participation from WillowTree Ventures and MVP Ventures, as well as several Fortune 100 operators who joined as angel investors and advisors. The funding brings Fibr AI’s total capital raised to $7.5 million.

For large enterprises, the disconnect between highly personalised ads and largely generic website experiences has traditionally been addressed through a combination of personalisation software, in-house engineering teams, and marketing agencies. This approach is costly, slow, and difficult to scale. While ads can be customised almost instantly, altering the experience a user sees after landing on a website often requires weeks of coordination and limits companies to running only a small number of experiments each year.

Fibr AI argues that this human-intensive operating model no longer makes sense. Instead, the startup deploys autonomous AI agents that infer visitor intent, generate variations, and continuously optimise webpages in real time.

Fibr AI replaces agency- and engineering-heavy workflows with autonomous systems that run continuously, said Ankur Goyal, the company’s co-founder and chief executive, in an interview with TechCrunch.

“We are the software, and the agency is the workforce of agents we are deploying,” Goyal said, adding that this enables Fibr AI to run thousands of experiments in parallel, rather than just a few dozen per year.

Early adoption was slow. Founded in early 2023 by Goyal and Pritam Roy, Fibr AI had only one or two customers for much of its first two years, as large enterprises took time to evaluate the approach. That began to change last year, Goyal said, as adoption increased among large U.S. organisations, including banks and healthcare providers, bringing the company’s total customer count to 12.

“We are an infrastructure afterthought layer,” Goyal said. “Once it’s set up, nobody wants to think about it again.” As a result, Fibr AI has secured three- to five-year contracts with large enterprises, which tend to treat website infrastructure as something to standardise rather than continually revisit.

Technically, Fibr AI functions as a layer on top of an existing website, integrating with a company’s advertising, analytics, and customer data systems to understand how visitors arrive and what they are likely seeking. Its AI agents then assemble and modify page elements — including copy, images, and layout — treating each URL as a dynamic system that learns and improves continuously, rather than as a fixed page.

Instead of relying on manually configured rules or sequential A/B testing, the platform runs numerous micro-experiments in parallel, updating experiences in real time as traffic flows in from different acquisition channels.

This approach has meaningful cost implications for enterprises. Traditional website personalisation typically combines software licensing fees with agency retainers and engineering costs, tying spending to headcount rather than results. Goyal said companies are increasingly evaluating Fibr AI based on price per experiment and conversion impact, rather than the number of tools or people involved.

For Accel, it was this operating model — rather than the hype around AI — that drove its decision to reinvest. “Advertising today is one-to-one, but when users land on a website it becomes one-to-many,” said Prayank Swaroop, a partner at Accel. “You can create hundreds of ads for different audiences, but they all still land on the same page.”

Fibr AI’s ability to remove agency and engineering bottlenecks and deliver one-to-one personalisation at scale stood out, Swaroop said, because those constraints typically limit how much experimentation enterprises can realistically pursue.

He added that early traction with banks and healthcare companies helped validate the model. “These are regulated, conservative industries,” Swaroop said. “When they start saying, ‘We need this, and we’re willing to pay for it,’ that’s when we feel confident doubling down.”

Preparing for the Agentic-Commerce Era

While most of Fibr AI’s current business focuses on personalising experiences for human visitors, both Accel and the startup see future potential in AI agents increasingly mediating online discovery. As users rely more on large language models and AI assistants — including OpenAI’s ChatGPT — to research and compare products before visiting websites, the ability for sites to adapt based on what a visitor or an AI agent already knows could become increasingly valuable.

“That part is still early,” Swaroop said, “but the companies building for today’s needs while being ready for that shift tomorrow are the ones we want to back.”

With the new funding, Fibr AI plans to expand its sales and customer-facing teams in the United States, while continuing to grow its technical operations in India. The San Francisco-based company maintains an office in Bengaluru, with 17 of its roughly 23 employees based in India and the remaining six in the U.S.

Goyal said the startup is targeting $5 million in annual recurring revenue by the end of this year and aims to serve around 50 enterprise customers.

Fibr AI is entering a market long dominated by incumbents such as Adobe and Optimizely, which offer experimentation and personalisation tools to large organisations. However, both Goyal and Swaroop argue that those platforms are constrained by how they are built and sold, often relying on agencies and internal engineering teams to configure and operate them.

That model, they said, makes it difficult to move quickly or scale experimentation, even as customer acquisition and messaging have become increasingly dynamic.

“Incumbents have been slow in bringing out products,” Swaroop said, adding that even when new features do arrive, they often appear years after market demand has shifted.