The situation
Logistics runs on documents, coordination, and high-volume communication. Across the industry, that still means people manually reading emails, extracting data, entering it into ERP systems, and replying one by one. Neelevat saw the opportunity to change that. Their operations team was spending hours each day on work that AI could handle — time that could go toward improving processes and building the business instead.
At the same time, the frontend their team relied on daily had reached its limits. Neelevat needed two things at once: AI-powered automation to reclaim operational capacity, and a ground-up rebuild of the tools their people used every day.
How we embedded
We didn't arrive as a separate team with a separate roadmap. For the AI workflow work, we sat inside Neelevat's operations — understanding how their team actually worked, where time disappeared, and which manual steps were candidates for automation. For the frontend, we stood up a dedicated Mino that integrated directly with the existing engineering team. Our UX/UI Lead joined their standups, collaborated with their developers, and shared ownership of the product from day one. Two tracks, one team.
What we built and why
Human-in-the-loop AI workflows, not full automation. On the operations side, we designed and deployed multiple AI-powered workflows targeting the most time-consuming manual processes — reading and extracting data from documents and emails, entering it into ERP, computing logistics responses, and drafting replies. The AI handles the heavy lifting, but people stay in control of decisions. This wasn't about replacing the team. It was about giving them their time back so they could focus on the work that actually requires human judgment.
A frontend built around the business, not a reskin of what existed. On the frontend, we started from scratch. We studied the business, mapped the user journeys, and designed a completely new UX before writing a line of code. The Mino took on full responsibility for building the interface, working alongside the existing team rather than in a silo. The result is a product built around how Neelevat's people actually work — not a reskin of what they'd outgrown.
The evidence
3,000+ hours saved in production. Real, measured hours that were previously spent on manual data entry, email processing, and operational back-and-forth. That time didn't just improve efficiency metrics. It reduced payment delays, accelerated customer response times, and improved satisfaction in ways that showed up in the business, not just on a dashboard.
More importantly, it freed Neelevat's team to think beyond firefighting. People who had been stuck in repetitive operational cycles could finally turn their attention to improving processes and shaping product direction.
What we're building together
We haven't left. We're still embedded — continuing to improve operational workflows and working with Neelevat's team on the frontend evolution. The AI systems are running in production. The new interface is taking shape.
The engineering culture we've built together — shared standups, shared ownership, human-in-the-loop as a design principle — is how the work gets done now. Not something we imposed and walked away from.