A phone number is one of modern life’s least glamorous objects. It is typed into checkout forms, recited to banks and copied into contact lists. Chirag Bakshi looked at the same string of digits and saw a small, live record of trust. Behind it sat a carrier account, a device, an activation date, a history of ports and SIM changes, and clues about whether the person holding the phone was the person an institution expected.
That observation became the organizing idea of Zumigo, the San Jose company Bakshi founded and still leads. In his telling, the phone number could act as a proxy for identity during a bank login or an online purchase. The useful part was not simply sending another one-time code. It was checking current information around the mobile account, often without asking a legitimate customer to stop, scan a document or answer a miniature census.
Bakshi’s route to that idea was long enough to resemble a history of networked computing. He worked on distributed PBX switch software at ROLM, helped lead engineering at Imagen, ran Apple’s LaserWriter product line, built a network graphics server business at Electronics for Imaging, co-founded the wireless platform company Unimobile and then led mobile messaging at VeriSign after it acquired the startup in 2004. Each chapter dealt with different machinery. Each asked a similar question: how does useful information move reliably through a network?
The career was the research
At ROLM, the network carried voices through office telephone systems. At Apple, a network made an expensive printer available to a room full of computers. At EFI, graphics servers managed the difficult trip from a digital file to a printed page. Unimobile and VeriSign moved messages between carriers, enterprises and people. By the time smartphones became a front door to commerce, Bakshi had spent decades working near the pipes.
This is why Zumigo’s founding insight feels less like a lightning strike and more like accumulated attention. Mobile accounts were becoming unusually durable pieces of a person’s digital life. Email addresses could be created quickly. IP addresses could be masked. A carrier relationship, by contrast, came with operational facts that were harder to improvise. Bakshi saw those facts as material for authentication.
“A person’s mobile phone number is a proxy to their identity.”Chirag Bakshi, on Zumigo’s founding idea
The first version of Zumigo was broader. When the company surfaced publicly around 2010, it offered hosted mobile messaging and location-based services. It raised an early $420,000 and pitched context-aware applications. Over time, the center of gravity shifted toward identity, transaction security and fraud prevention. The phone was still the point of entry, but location became one signal among many rather than the entire story.
- 1980sDistributed telephone software at ROLM, followed by engineering leadership at Imagen.
- 1990sApple’s LaserWriter line, then a network graphics server business at EFI.
- 2000-2008Unimobile founder and COO, then VeriSign mobile messaging leader after the acquisition.
- 2009 onwardZumigo moves from mobile context and location toward identity, authentication and risk intelligence.
A quiet security product
Security software has a habit of announcing itself through inconvenience. It asks for another password, another code, another document, another wait. Bakshi’s product philosophy runs in the opposite direction. He says fraud prevention and user experience must move together. When the network can establish confidence in the background, the honest user should not have to perform more work merely because a dishonest user exists.
The approach starts by reading the mobile account as a bundle of evidence. Is the number active? How long has the account existed? Is it prepaid? Was the number recently ported? Did its SIM or device information change? None of those details proves identity alone. Layered together, and combined with email, payment, device or location information, they can produce a more useful risk assessment.
The identity hiding around a phone number
The distinction between a signal and a verdict matters. A recent SIM change may be entirely ordinary. A customer may have upgraded a phone that morning. The product decision is not simply whether to block. It is whether to let the transaction pass, ask for another factor or alert the institution that the context deserves attention. Good risk systems make room for uncertainty without turning every uncertainty into punishment.
Zumigo also uses verified mobile information to prefill forms. This places conversion and security on the same side of the ledger: fewer fields for a customer to type, and better evidence behind the fields that appear. In 2014, Bakshi and colleague David Pinski demonstrated that idea at FinovateSpring. The product matched names and addresses, evaluated the device and used geolocation as part of a three-part check. The demo made an invisible carrier relationship legible to a merchant.
A low-friction decision path
From one account to a network
Bakshi’s more recent idea expands the frame. Fraud rings do not confine themselves to one bank or one retailer. Stolen identity details can be used across many institutions in quick succession. A phone number that looks merely unusual inside a single company may look plainly coordinated when its activity is visible across a broader network.
He calls the response “organized fraud prevention.” The model uses activity from banks, fintechs, neobanks, insurers and merchants to detect patterns that no single participant can see as quickly on its own. Zumigo’s Assure Insights product turns device-level and account-level activity into a trust score, giving businesses a reason to change the next authentication step before a suspicious attempt becomes a completed loss.
In January 2026, the product received a Gold award for real-time fraud intelligence sharing in Juniper Research’s Telco Innovation Awards. Bakshi used the occasion to emphasize speed: fraud can occur in milliseconds, he said, so current data is an advantage. The award fits the long arc of his career. The network is no longer just carrying a voice, a print job or a message. It is helping participants understand an event while there is still time to act.
Scale does not settle the harder questions of digital identity. It makes judgment more important. A system should gather enough context to separate routine behavior from coordinated abuse, but the person on the other side still wants a simple login and a reasonable transaction. Bakshi’s answer is modularity: use more layers when the stakes are high, fewer when someone is signing up for something trivial.
The password afterlife
Bakshi has repeatedly predicted that enterprises will move away from passwords. The reasons are familiar: stolen credentials, phishing, password reuse and now AI-assisted attacks that make impersonation cheaper. Passkeys and SIM-based authentication can remove some of the weakest rituals. Yet passwordless does not mean identity-less. Before a device receives a passkey, someone still needs confidence that the person installing it owns both the account and the device.
That gap is where Bakshi keeps returning to carrier information. In 2025, Zumigo announced SIM-based authentication with major US network operators, along with a visual low-code tool for companies to assemble their own verification workflows. It also added device fingerprinting and a broader phone trust score. The practical ambition was to make the machinery easier for a business to deploy while keeping it nearly invisible to a legitimate customer.
“User experience and fraud prevention must go hand-in-hand with any identity proofing solution.”Chirag Bakshi
Bakshi is an engineer by training, with a B.Tech. from IIT Bombay and master’s degrees from Stanford and the University of Michigan. His name appears on patents covering location alerts, mobile number verification, device-history risk indicators and network-based authentication. But his public language is usually less about a clever mechanism than the operating result: current evidence, fewer false alarms and less friction.
There is a small clue to his personality in Zumigo’s annual prediction posts. Alongside forecasts about passwords and artificial intelligence, he predicted that the San Francisco 49ers would reach the 2024 Super Bowl. They did, losing in overtime. The football call adds a little play to an otherwise serious body of work. It also resembles his professional habit: read the available signals, make a judgment before the result is known and accept that confidence is never certainty.
What the long view reveals
Founder stories are often compressed into a moment of inspiration. Bakshi’s is more useful when left uncompressed. The identity thesis drew on telephone systems, network printing, mobile platforms, carrier relationships and years of watching people move more of their lives onto a device. It was a startup idea produced by continuity.
The pattern offers something worth stealing. Stay close to a consequential layer of infrastructure. Watch what moves across it. Notice the new responsibilities it acquires as technology changes. A printer network can teach scale. A messaging network can teach interoperability. A mobile account can become part of an identity system once commerce arrives in every pocket.
Bakshi now talks about autonomous identity verification: intelligent systems finding the connective tissue among billions of data points, while legitimate users move through the front door without a scene. The phrase sounds futuristic. The method underneath it is the one he has followed for decades. Look carefully at the quiet system everyone depends on. Find the signal hiding there. Then make it useful before the moment passes.