InstaLILY closes $60M Series B for physical-goods AI agents

The forward-deployed engineer model is quietly becoming a product, not just a hiring trick. InstaLILY, which builds domain-specific AI 'teammates' for physical-goods and services firms, closed a $60 million Series B according to SiliconANGLE, bringing its total raised to almost $100 million. Energize Capital led the round, with previous backer Insight Partners and new investors Home Depot Ventures and United Rentals participating.

The pitch is narrower and stranger than the usual horizontal-copilot story. InstaLILY builds 'InstaWorkers,' autonomous agents that live inside a distributor's or industrial firm's workflows and automate sales, operations, and service tasks. Named customers include SunSource, Parts Town, SRS Distribution, United Rentals, ShipStation Global, Henry Schein, and PartsSource, spanning construction, logistics, healthcare, and field service. That reads less like a Silicon Valley launch list and more like the guest book at an industrial distribution conference. Alongside the round the company launched Lily, described as an 'AI forward-deployed engineer' trained on a customer's business processes to develop custom software applications. CEO Amit Shah's line for it is that 'Lily turns that work into software, and it is running in days, not quarters.'

The claim that anchors the round is a revenue jump of more than fivefold over the past year, and a single national distributor that reportedly identified new sales opportunities generating over $200 million in new revenue. Both are the company's telling, not audited numbers, so take the specifics as reported, not settled. The strategic-investor mix is the more revealing signal: when Home Depot Ventures and United Rentals put money into an AI vendor whose product they can also buy, it usually means the automation is already inside the tent, and the corporate wants first-look and pricing leverage as much as financial upside.

What the reporting doesn't give you is the valuation, the pricing model, or how much of that fivefold jump is software versus forward-deployed services, which is the difference between a durable software business and a very well-funded consultancy with an agent runtime attached. The bet worth watching is whether vertical AI plays like this one keep quietly outrunning the horizontal copilots in industries where the workflows are ugly enough that no generalist model can just walk in.