Midas, a NYC-based company that provides verification layer services, has raised $10 million in funding, led by Valour Equity Partners and Nova Global.
Read More:Alcove Raises $1M in Pre-Seed Funding
The company plans to use the funds to expand its operations and further develop its products.
Midas is a company that uses mathematical verification to make AI systems more reliable and secure. The team includes 11 medalists from the International Mathematical Olympiad (IMO) and the International Olympiad in Informatics (IOI), two of the world’s most selective academic competitions. They bring experience from companies such as Jane Street, Google, AWS, NVIDIA, and Mercor, as well as educational backgrounds from Stanford, MIT, Cambridge, Princeton, and Duke.
The company is supported by investors who backed OpenAI, Tesla, and SpaceX, showing early confidence from people experienced in building some of the world’s most important technology companies.
Together, they are using formal mathematical verification to tackle one of AI’s most significant challenges: building trust.
“Shalim Monteagudo-Contreras, President and Co-Founder of Midas, said, ‘Modern AI can generate convincing answers, but it can’t prove they are correct. Midas creates a barrier between AI’s probabilistic outputs and real world systems, enforcing correctness mathematically so that results are proven, not guessed, argued, or hoped for, before being used.”
Fluency can’t be audited—but proof can.
Renzo Balcazar, CEO and Co-Founder of Midas, said, ‘Every human institution—from law to science to finance—runs on evidence. Artificial intelligence is the first form of intelligence that operates without it.’ Unlike humans, AI is deployed at scale without a proof loop,
producing results without explanation, confidence without causality, and output without evidence. As AI systems generate answers faster than humans can verify, coherence is often mistaken for correctness, fluency replaces evidence, and confidence takes the place of truth.
Plausibility can scale, but proof cannot—unless it’s built into the system.
Rodrigo Porto, Tech Lead at Midas, said that building verification into the reasoning process from the start—rather than checking for errors afterwards—is what makes trust possible as AI systems become too complex for manual review. Midas embeds mathematical proof at the core of AI, verifying outputs, data, and reasoning so these systems can be trusted in situations where mistakes are unacceptable.
“John Stanton, Vice President at Valour Equity Partners, said, ‘Every layer of the AI ecosystem eventually needs its own infrastructure, and verification is the final missing piece. This isn’t about probabilities—it’s about proof. What makes Midas different is its culture: the team rejects ‘almost correct’ answers and accepts only what can be proven.
“Carlo Agostinelli, founder of Nova Global, said, ‘At Nova Global, we back founders with the potential to make history. Shalim Monteagudo-Contreras and Renzo Balcazar are already at that level. They’ve built a world-class team from scratch and are tackling one of AI’s most fundamental challenges: trust. Their proof-first approach to AI reliability shows both technical ambition and perfect founder-market fit, making Midas a generational company.’
About Midas
Founded by CEO Renzo Balcazar and President Shalim Monteagudo-Contreras, Midas is creating the verification layer for AI. This mathematical trust system uses formal verification to ensure AI outputs and training data are provably correct.
The team includes 10 IMO and IOI medalists from Cambridge, MIT, Princeton, Duke, and Stanford, as well as senior engineers from top tech companies. Midas applies formal mathematics to safely deploy AI across critical sectors such as biotech, defence, hardware design, and finance: Valour Equity Partners, Nova Global, and other leading investors back the company.
Read More:Uptool Raises $6M in Seed Funding
