A career built on inconvenient truths
Today Aran Khanna runs Archera, a Bellevue startup that promises cloud buyers a deal that sounds too good to audit: deep discounts on AWS, Azure, and Google Cloud without the multi-year handcuffs those discounts usually demand.
The trick is insurance. Big cloud providers reward commitment - sign up for one or three years and the per-hour price drops. The catch is the bet: commit to capacity you might not use and you eat the waste. Archera steps into that gap, letting companies take short-term flexible commitments while standing behind the risk with reinsurance and lending capacity. In July 2024 the company closed a $17 million Series B led by HighSage Ventures, alongside Ridge Ventures, Amplify Partners, and PSL, plus access to more than $100 million in reinsurance and lending capacity. The 22-person company said it had grown its revenue run rate 500% year over year and crossed 400 customers.
Khanna co-founded Archera in 2019 with his brother, Nikhil. The idea came from inside the machine. As an engineer on the early SageMaker team at Amazon Web Services, he was handed a hard job: drive customer adoption of a very expensive GPU compute product. He watched the largest enterprises automate cloud management into a science while smaller companies did the same work by hand, burning money and engineering hours. Archera is his answer to that asymmetry.
"How can we be most aligned with the customers? How do we have a model that's as customer-obsessed as possible?"
The extension that named everyone's hiding place
Before the cloud, there was the map. In May 2015, as a Harvard senior studying computer science jointly with mathematics, Khanna wrote a Chrome extension and called it Marauder's Map, after the Harry Potter parchment that reveals where every person in the castle is standing. His version did the same to Facebook Messenger. By reading the location data that Messenger quietly attached to messages, the extension plotted on a map exactly where people were when they typed - including strangers sharing a group chat.
He could find an acquaintance's dorm room, he explained, by looking at the cluster of messages sent late at night. That single detail is what made the abstract land. Privacy lectures are easy to nod along to and forget. A pin dropped on someone's bed at 2 a.m. is not.
"We are constantly being told how we are losing privacy with the increasing digitization of our lives, however the consequences never seem tangible."
He posted about it on Twitter, Reddit, and Medium. It spread fast - more than 85,000 downloads. Facebook pushed an update to Messenger to throttle the location sharing. Three days after he went public, Facebook withdrew the summer internship offer it had extended him. The story ran in the Washington Post, on CBS, and across the tech press, usually framed as a cautionary tale about hacker culture meeting corporate caution.
Khanna's read was different. He said he had no malicious intent and no regrets. The point was never the dorm rooms. It was that the data was already being shared, by default, by everyone - and almost nobody could see it.
Then he did it again, with money
Later in 2015 he turned the same lens on Venmo. Working through Harvard's Institute for Quantitative Social Science, where he spent 2015 to 2016 as a research fellow and developer, he built a tool called Money Trail and published research showing how the public-by-default feed of roughly 350,000 Venmo users exposed their financial lives and social graphs. Payments that feel innocent in isolation - rent, drugs jokes, breakups rendered as emoji - become, in aggregate, a way to identify, classify, and incriminate people. Years later, when financial privacy finally became front-page news, he noted on X that he had documented every one of those issues back in 2015.
From gadfly to builder
The privacy work was not a phase he abandoned so much as a way of thinking he carried into engineering. Before Archera he interned at Microsoft, at the hedge-fund data startup Novus, and at the deep-learning shop Marianas Labs. In 2016 he joined AWS as an AI engineer, co-authoring academic papers and helping build neural-network compression tools - work that shows up in his Google Scholar profile and in patents related to privacy-aware machine learning. From 2018 to 2019 he co-founded Glia Intelligence, a machine-learning consultancy optimizing retail replenishment and pricing with deep-learning models.
The throughline is consistent. Khanna is interested in what happens to data once it leaves your hands - whether that data is your location, your transactions, or your company's cloud spend. The first two he exposed. The third he is trying to tame.
"With this code you can see for yourself the potentially invasive usage of the information you share, and decide for yourself if this is something you should worry about."
Watch & listen
Khanna is a TED speaker and a regular on engineering and cloud-economics podcasts. A good starting point is his talk "Putting Our Heads in the Clouds with Aran Khanna" on YouTube, where the privacy researcher and the cloud-finance founder turn out to be the same person asking the same question from two directions.