Dhruv Jaglan on Why 96% of Businesses Still Automate Nothing

Enterprise software has never been more capable. Companies have access to powerful AI tools, no-code platforms, and workflow automation products that promise to eliminate repetitive work and free up teams for higher-value tasks. Yet the results, in most organizations, tell a different story. A recent study by MIT Sloan concluded that "The 95% failure rate for enterprise AI solutions represents the clearest manifestation of the GenAI Divide," pointing to a persistent gap between the technology's potential and what organizations are actually able to capture. For most companies, automation remains something that happens in engineering, not across the business.

The Bottleneck Nobody Has Solved

"The business case is clear. What to automate is obvious," said Dhruv Jaglan, Founder and CEO of Rilo, a workflow automation platform designed for knowledge workers. "The problem is that every tool on the market still requires you to think like an engineer. And the people who most need automation are not engineers."

It is a bottleneck that has persisted across multiple generations of automation technology. Rule-based platforms gave way to visual drag-and-drop interfaces, which gave way to low-code environments, yet the underlying requirement remained: users needed to understand triggers, conditional logic, data transformations, and APIs. For non-technical teams in sales, marketing, operations, and customer success, this learning curve proved consistently prohibitive. According to Formstack's digital maturity research, "Only 4% of businesses have achieved a digitized, fully automated workspace," a figure that has barely moved despite years of investment in automation tooling.

Building AI Before the AI Era

Long before workflow automation became a mainstream conversation, Dhruv Jaglan was working on a quieter version of the same problem: how do you build AI systems that people without technical backgrounds can actually trust with consequential work? His early academic work, including research in medical imaging at the Technical University of Braunschweig, pointed toward an enduring interest—not just in building capable systems, but in applying them where reliability is non-negotiable.

The venture that sharpened that instinct most was Babblebots, the AI recruitment platform Mr. Jaglan co-founded. The core insight was simple but technically demanding: large enterprise hiring teams were drowning in applicants they could not meaningfully evaluate. A company with ten thousand unscreened candidates does not have a shortage of talent. It has a shortage of bandwidth. Babblebots' AI interviewers worked through that backlog, conducting voice assessments across more than twenty languages simultaneously, surfacing the outlier candidates that manual screening would have missed entirely. The result was not just faster hiring but better hiring. Companies were finding stronger fits from the pool they already had, without increasing headcount in their recruiting teams. What made this technically significant was the moment in which it was built: before ChatGPT existed, when most AI systems could not reliably handle basic language tasks, let alone the nuance of a multi-lingual, high-stakes candidate conversation.

"The people running those hiring processes were not lacking judgment," Mr. Jaglan said. "They knew exactly what a good candidate looked like. What they could not do was build the technology to find that candidate at scale. That disconnect between knowing what you need and being able to build for it is not unique to recruiting. It shows up in every function across every company."

Babblebots compressed hiring timelines from three or more weeks to a matter of hours, and was adopted by more than twenty-five companies across industries ranging from publicly listed enterprises to early-stage startups. The platform's voice AI layer detected nuance, asked follow-up questions, and adapted dynamically to candidate responses, capabilities that would have seemed implausible from a pre-GPT system to most observers. Mr. Jaglan's early research at the Technical University of Braunschweig on 3D reconstruction algorithms for medical applications, which resulted in a peer-reviewed publication, had already demonstrated his ability to work at the technical frontier before the tools caught up with the ambition.

Rethinking Who Gets to Automate

Rilo is Mr. Jaglan's attempt to carry the lessons of Babblebots into a broader problem. Where Babblebots made sophisticated AI accessible to HR professionals who lacked technical backgrounds, Rilo is designed to do the same for the full range of knowledge workers who understand their own workflows but cannot translate that understanding into automation.

Users describe what they want in plain language, the same way they would explain a task to a colleague, and Rilo converts that description into a production-ready, multi-step workflow that runs across their existing tools. The platform integrates with over one hundred business applications, including Salesforce, HubSpot, Slack, Gmail, Notion, and LinkedIn, and selects automatically between multiple AI models to handle each step of a workflow, ensuring quality without locking users into any single provider.

"We are not building another tool that requires you to think programmatically," Mr. Jaglan said. "We are building a platform where knowing what to automate is enough to get it done. No code. No technical bottleneck. Just describe it."

The platform also incorporates browser-based visual automation, enabling it to operate within web applications that do not offer API access. This extends automation reach into legacy enterprise systems, custom internal tools, and older platforms that have historically been impossible to automate through conventional means.

Backing and Early Traction

Rilo has raised $1.5 million in seed funding through Peak XV Partners (formerly Sequoia Capital India) Surge—"a seed platform for company builders" combining capital with hands-on company-building support—with additional participation from De VC, Day Zero Ventures, and investor Gautam Prakash. The company is initially focused on sales and marketing workflows, where the volume of repetitive tasks is high and the consequences of automation errors are manageable, before expanding into higher-stakes operational functions.

Early adopters report results consistent with what Mr. Jaglan observed at Babblebots: meaningful time recovered, operational processes running without manual intervention, and teams reorienting their energy toward work that actually requires human judgment. One founder reported month-over-month growth of 25% that he attributed directly to Marketing automations built on the platform. Another described the system taking over post-call CRM updates entirely, eliminating a category of administrative work that had consumed significant sales team capacity.

The Gap That Remains

Mr. Jaglan is measured about what remains unsolved. The automation market is competitive, and the promise of effortless workflow automation has been made before without delivering at scale. What he believes differentiates Rilo is not the breadth of integrations or the sophistication of the underlying models, but the design philosophy: building for the person who understands the process, not the person who can configure the technology.

"Most automation tools are built for technical users and then retrofitted for everyone else," he said. "We started from the assumption that the person who knows the most about a workflow is almost never the person who can automate it. That assumption changes everything about how you design the product."

For the 96% of businesses that have not yet fully automated their core workflows, that design philosophy may prove to be the difference between another failed pilot and an automation program that actually scales. Mr. Jaglan has spent his career building AI systems that non-technical users could trust with consequential tasks. Rilo is the most ambitious version of that work yet.