Off-the-shelf cameras on drones and forklifts turn a warehouse into a live inventory map — sold as an outcome, not hardware, at ~$500K a year per customer
largest customer $3–4M/year · 170% net revenue retention · 2.5× growth last year, 2–3× forecast this year · $74M raised to date including a $40M Series B led by Smith Point Capital · ~75 people
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Sankalp Arora did a PhD in robotics at Carnegie Mellon, where he and his two cofounders, Daniel and Gitesh, built what he describes as one of the world's first fully autonomous helicopters, for DARPA — work that won national awards. His thesis was about making robots curious. In a warehouse, curious means about boxes, barcodes, inventory and workflows. The first line of code for Gather AI was written right after his PhD defence in late 2018. Then came three years of nothing but building. He is direct about why: nobody had ever built this stack around hardware you can buy retail, and they refused to ship until it performed reliably in live customer environments. First customer landed in 2021. To fund those three pre-revenue years they raised, on the explicit reasoning that they were pitching technology that did not exist and therefore needed real capital. Since then: 2.5× revenue growth last year, forecasting 2–3× this year, 170% annualised net revenue retention, 30–40 customer logos including Geodis, Axon and Barrett Distribution Centers, and roughly $10–15M of run rate. The company has raised $74M in total, most recently a $40M Series B led by Smith Point Capital — the fund founded by Keith Block, the former Salesforce co-CEO, with Chris Little. The trigger for that round was not a technology milestone but a sales one: the motion had scaled beyond founder-led selling, and Arora notes with evident relief that the new sellers were outperforming him. He also volunteers, unprompted, that he lived with suicidal thoughts from his teens until about eighteen months before the interview — through the period the company passed $9M of revenue — and that finding the right treatment changed how he operates. That is not incidental colour; it is the most useful thing in the interview for anyone considering this path.
6 more sections — the full story, the playbook, the risks, and the numbers. Free account unlocks everything.
Create free accountEmail code only. No password.The quick-reference card — acquisition channels, replication playbook, and risk map — unlocks with a free account.
Create free accountEmail code only. No password.Paste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea: a full build spec and an SEO growth plan. Copy, paste, ship.
Go Pro — $10/moCancel anytime.Every claim traces back to a source video.
Source video links unlock with a free account. Full transcripts are Pro.
Create free accountEmail code only. No password.Data credibility: Founder interview on a promotional SaaS podcast; the host is also an investor soliciting deal flow during the episode. All figures are founder-stated and unaudited: ~$500K average annual deal size for 5–7 facilities, $3–4M for the largest customer, 30–40 customer logos, 170% annualised net revenue retention, 2.5× growth last year and 2–3× forecast, $74M raised to date, a $40M Series B led by Smith Point Capital (Keith Block and Chris Little), ~75 employees, and ~150,000 US warehouses over 100,000 square feet. The headline run rate is not a founder disclosure — the host multiplied 30 logos by $500K to get $10–15M and the founder replied 'roughly right', so treat it as an accepted estimate rather than a stated number. The valuation range of $270–400M is the host's inference from assuming 10–15% dilution on the Series B, not a disclosed figure. Two transcript defects matter: the largest-customer number is rendered once as 'three to four billion dollars a year' before both parties settle on millions, and proper nouns are mangled throughout ('Sankalp Aura' and 'SunRob' for Sankalp Arora, 'Autobytel' and 'MHA Vision' for MHE Vision, 'SeaGrid', 'Geodis'). The Barrett Distribution Centers / Stadium Goods case — half a million pairs of sneakers, $2 picks escalating to $10–15, a 70% reduction in location errors — is described by the founder, not independently verified. No dashboard was shown and no third-party data was cited.