Deepak Aatresh Is Teaching Buildings to Think Like Chips

In 1983, a battery-powered car rolled onto the roads of Bangalore with a research engineer at the controls and his teenage son along for the ride. Reporters followed. Deepak Aatresh watched his father test something India was not yet prepared to back. The prototype worked, but the project stopped at the prototype. Money and institutional confidence never arrived.

The episode gave Aatresh two early lessons that would surface again in his own career. A technical idea needs more than technical merit. And a system can prevent an invention from becoming a product even when the invention itself is sound. He wanted to follow his father into electrical engineering, but he wanted to do it somewhere enthusiasm for technology could turn into momentum.

A magazine helped draw the route. Aatresh read about Intel's effort to build its first million-transistor chip, studied electrical engineering in Bangalore, then left India in 1987 for Arizona State University. He completed a master's degree in electrical and computer engineering. Intel recruited him to Santa Clara in 1989. He later joked that Silicon Valley fever took about a day to set in.

The discipline of tiny things

At Intel, Aatresh spent seven years working on the architecture and design of successive generations of microprocessors. A chip is small enough to sit on a fingertip, yet its design is a negotiation among huge numbers of components, constraints, and trade-offs. Human drafting alone cannot hold the whole problem. Semiconductor teams learned to rely on reusable libraries, formal rules, automated layout, simulation, and verification before manufacturing.

That environment taught a specific kind of confidence. Complexity does not always need to be simplified away. It can be described precisely enough for a machine to explore. Errors caught in software are less expensive than errors found after fabrication. A functional intention can be turned into rules, and rules can produce a physical spatial arrangement.

After Intel, Aatresh moved through communications hardware. His public career record includes Cabletron Systems, Riverstone Networks, and Alcatel-Lucent. One of his early inventions, filed while at Cabletron, concerned how packets competing in multiple queues are forwarded to a finite-bandwidth output. The object was network traffic, but the habit was already there: define the constraints, arbitrate competing needs, and make the flow work.

A year of construction, compressed

In 2008, Aatresh joined Artiman Ventures as an entrepreneur in residence. He had the startup impulse before he had the startup idea. Energy waste interested him, and buildings were an obvious place to look. The difficulty was equally obvious: he did not know the construction industry.

So he studied it. He traced the sequence from site clearing and excavation through the rise of a structure. During that research, he watched a fixed camera compress more than a year on a construction site into a few minutes. Ground was removed. Materials arrived. Layers accumulated. A process that looked slow and fragmented at human speed suddenly resembled fabrication.

“That’s when it hit me. It’s just like we make silicon chips, first we etch away, and then we deposit layers.”Deepak Aatresh

The analogy did not mean a building was literally a chip. It meant two complicated physical objects could share a design philosophy. Semiconductor engineers no longer place every element by hand. They describe behavior, draw from component libraries, test constraints, and let software manage combinations at a scale people cannot comfortably enumerate.

Aatresh started airing the idea widely. Skeptics pointed out that chips are replicated while every building is different. His answer was practical: chips differ from one another too. Repetition was not the only source of automation. Rules and patterns could accommodate variation.

Finding the other half of the language

In 2009, Aatresh connected with architect Zigmund Rubel. The pairing supplied the company with necessary tension. A chip designer could bring automation and verification, but architecture carried codes, materials, workflows, aesthetics, and the lived realities of space. Aitresh and Rubel founded Aditazz in 2010 with seed backing from Artiman Ventures. The name was adapted from a Sanskrit expression for starting from the beginning.

Their system would describe recurring spatial patterns and the rules that connect them. A client could state functional needs and constraints. Software could assemble, adjust, and compare layouts. A building became something closer to a compiled outcome: an arrangement generated from a functional description, then evaluated before expensive commitments hardened into concrete and steel.

The founding pair also had to decide what kind of business could prove the premise. Aatresh's instinct was software. Yet a new design engine could not simply arrive with a login screen and expect a conservative industry to trust it. Aditazz first acted as a service provider, using real commissions to test its algorithms and establish that the output could survive professional scrutiny.

This is the less glamorous part of cross-industry invention. The analogy may arrive in a flash, but adoption moves by demonstration. The software had to meet architects, engineers, contractors, owners, and regulators where they already worked. A later industry discussion associated with Aatresh distilled the operating rule: no matter how complex the concept, implementation must feel easy.

The useful chaos of being early

An early proving ground came through Kaiser Permanente's Small Hospital, Big Idea competition. About 300 firms applied. Aditazz became one of two winners. The young company received an initial award, invested it in prototypes, and then used those prototypes to produce its final submission.

Validation came bundled with administrative comedy. Aatresh recalled the organizers asking for two years of tax returns. Aditazz had not existed for two years. The moment captures the startup's position neatly: its work was credible enough to win, while the organization itself was still younger than the paperwork expected of it.

The competition gave Aditazz more than a result. It gave the software a demanding, legible test. Requirements could be formalized. Layouts could be compared. Trade-offs could be surfaced before construction. The company went on to apply the approach to other complex facilities and to patent methods for matching form to function and evaluating building configurations.

What stayed constant

Aatresh's career can look like a large jump from microprocessors to architecture. Up close, the continuity is stronger. At Intel he worked with components competing for space, power, and performance. In networking, queues competed for finite bandwidth. At Aditazz, rooms, systems, circulation, cost, and geometry compete inside a finite envelope. The scale changes. The engineering question survives.

He also kept the immigrant's double vision. A 2015 visit by India's prime minister to Silicon Valley prompted Aatresh to talk publicly about returning knowledge and investment to the country he had left. By then Aditazz had opened an office in India. He was candid about the friction of doing business there and equally direct about the pull. “In the end, it's home,” he said.

The personal connection was present in smaller ways too. A profile from that period found him starting the day in Saratoga with South Indian filter coffee. It is a modest image beside the abstractions of algorithms and digital models: a morning ritual carried across an ocean, while the work carried methods from one profession into another.

At a U.S. Army Corps of Engineers innovation gathering in 2019, Aatresh came away interested in problems Silicon Valley often overlooked and in the Corps' openness to private collaboration. The reaction fits the pattern. He tends to approach a field by looking for constraint, unmet capacity, and a possible translation.

That is the durable idea inside Aditazz. The company is not simply a story about putting computers into construction. It is about moving a mature way of thinking across a professional border. Describe the intent. Encode the rules. Generate alternatives. Test before committing. Then give experts a way to see and shape the result.

Aatresh began with a father's unfinished electric car, an article about a million transistors, and a year of construction reduced to minutes. Each moment showed a system in motion. His work since has asked what happens when those systems can borrow one another's grammar.