Not chatbots. Autonomous agents that read incoming work, reason over your rules, draft the output and execute after a human approves.
WorkAgents AI started with a simple observation: back-office teams in manufacturing, pharma and trading spend most of their day on repetitive, rule-bound work — reading RFQs, matching catalogues, sourcing vendor prices, drafting quotes, reconciling invoices. It's exactly the work AI is good at, if it's held to a strict human-in-the-loop standard.
So we built a platform where AI employees own whole workflows end-to-end, but never make a customer-facing or financial decision without an approver. Every action is drafted, reviewed and written to a tamper-evident audit trail. Incubated at Somaiya RIIDL, Mumbai, we're building the AI workforce for real trade.
Rohan founded WorkAgents AI with the mission of building AI employees for modern businesses. He leads product and engineering, working directly with early customers in manufacturing, pharma and trading to shape agents around real back-office workflows rather than demos.
WorkAgents AI is incubated at Somaiya RIIDL, the research, innovation and incubation centre at Somaiya Vidyavihar, Mumbai.
Draft-first AI. Agents never take a customer-facing or financial action on their own. Every quote, email and order is a draft until a named approver signs off — and each human edit locks the AI out of silently changing it again.
Auditability. Every action, whether performed by a person or an agent, is written to an audit log with actor, time and context, so you can always answer "who did this and why".
Honest economics. We don't promise magic. Automating rule-bound back-office work with review-and-approve AI typically costs a fraction of doing the same work manually — and frees your team for the judgement calls that actually need them.