CipherSonic AI: Funding Rounds, Investors & CEO

About CipherSonic AI

CipherSonic AI helps businesses use artificial intelligence on sensitive information without revealing the underlying data. It uses fully homomorphic encryption, a method that keeps information encrypted while a system processes it. It targets regulated settings such as healthcare and financial services. Its five products include CipherChat, CipherCloud and CipherInspect.

The company is based in Cambridge, Massachusetts, with another location in Bengaluru, India. Ajay Joshi is chief executive officer and co-founder. Rashmi Agrawal is chief technology officer and co-founder.

Funding Rounds (SEC Form D)

First sale	Filed with SEC	Amount sold	Investors	Avg. check	Terms	Filing
Jan 8, 2024	Nov 12, 2026amended Nov 13, 2026	$540,000fully sold	11	$49,091	Simple Agreement For Future Equity (SAFE)Rule 506(b)	Form D/A ↗1 earlier filing

First sale is the date the first investor in the offering committed to invest. Companies must file Form D within 15 days of that date, so filing dates run later.

Why Investors Backed CipherSonic AI

Businesses adopting AI need to protect training data, user inputs and the models themselves. Agrawal's research removed a major performance bottleneck in fully homomorphic encryption, making it practical. That breakthrough became the technical core of CipherSonic. [4]

Origin and Team

Ajay Joshi and Rashmi Agrawal co-founded CipherSonic in January 2024. It grew from Agrawal's Boston University research under Joshi on processing encrypted data. [4]

Joshi is a Boston University professor. He earned his doctorate from Georgia Tech in 2006, then conducted postdoctoral research at MIT. He also worked at Google and Lightmatter. His honors include the NSF CAREER Award and BU Ignition Award. [4]

Agrawal earned a Ph.D. in computer engineering from Boston University. Her doctoral work created programmable-chip accelerators for fully homomorphic encryption. [4]

Product and Technology

Fully homomorphic encryption lets systems compute with data while it remains encrypted. CipherSonic uses it so AI can operate without exposing user information or model intellectual property. [4]

The research uses programmable chips, called FPGAs, to speed encrypted cloud computation. It achieved more than a 1,000-fold speedup. Designs include the RACE RISC-V system-on-chip and FAB accelerator architecture. CipherSonic's cloud infrastructure is designed to resist future quantum attacks. [4]

Milestones

Scroll table to see all columns →

Date	What happened	Source	Coverage
2025	CipherSonic Labs received a Phase I SBIR award for work bridging multiparty computation and fully homomorphic encryption.	[5]	—
2025	CipherSonic was a winner or finalist for the MassChallenge RESOLVE '25 accelerator.	[4]	—
2025	CipherSonic was a winner or finalist for the BU Ignition Award.	[4]	—

Market

Joshi expects privacy to matter more as organizations depend on data. He sees encrypted collaboration supporting customer service and business insight without exposing customer information. [4]

We wrote this summary from public statements by CipherSonic AI and its investors. It is not investment advice.