NUMOS AI

Parijat's background is a study in the specific kind of cross-disciplinary depth that tends to produce good enterprise software founders. He was born in India and grew up in Abu Dhabi, before studying at the University of Pennsylvania and Wharton, which is where his engineering and business foundations formed simultaneously. The Penn/Wharton combination is significant: Penn Engineering produces rigorous computer scientists, while Wharton produces people who think in terms of organisational systems and financial incentives. Parijat absorbed both. He graduated in 2011 with a Bachelor of Engineering, and his early career took him through Microsoft, where he worked in product, before he joined Zenefits. Zenefits at the time was one of Silicon Valley's most scrutinised HR software companies, having gone through a dramatic growth-and-governance crisis before stabilising under new leadership. Parijat rose to Senior Vice President holding leadership responsibilities across product, growth, and engineering simultaneously, a scope that is unusually broad for someone who hadn't yet founded a company, and that reflects both his versatility and the confidence the Zenefits leadership placed in him. After Zenefits he spent time at South Park Commons, the San Francisco founder community and fellowship programme, and at TriNet before going into stealth with Numos. The framing he uses publicly, "finance teams aren't just analyzing numbers, they're accountable for them", is a precise articulation of why auditability is the product's core differentiator, and it is the intellectual entry point for any cold outreach that wants to resonate.

Mitul is one of the more academically and technically decorated CTOs at a seed-stage AI company right now, and his credentials are directly relevant rather than incidentally impressive. He completed his undergraduate degree in Computer Science and Engineering at the Indian Institute of Technology, Bombay, in 2001, ranked 49th among over 100,000 candidates in the IIT-JEE entrance examination that year, placing him in the top 0.05% of all examinees. He also placed in the top 25 students selected for the gold medal of the National Science Enrichment Programme in Physics in 1997. Those early academic signals established a pattern that continued throughout his career: he is someone who operates at the very top of whatever field he enters. He then moved to the University of Texas at Austin for graduate study, receiving both an NSF grant and a Texas Advanced Technology Program grant while completing his PhD in Computer Science in 2007, which focused on distributed systems, caching, and resource scheduling. His co-authored work during this period spans conferences including SPAA and IPDPS. After UT Austin he joined Kosmix, a web-scale text categorisation and entity extraction company that was later acquired by Walmart (becoming Walmart Labs), where he worked on large-scale information retrieval. He then spent years at LinkedIn, where he was Head of People You May Know and Growth Relevance, two of LinkedIn's most critical recommendation systems, each operating at hundreds of millions of users. His published work from this period includes papers at KDD, WWW, RecSys, SIGIR, and CIKM on social recommender systems, entity extraction, and collaborative filtering. He then co-founded Passage AI, a conversational AI company that was acquired by ServiceNow, after which he joined ServiceNow as Director of AI and Machine Learning Engineering, leading their natural language processing group and shipping production features including Conversational AI, Incident Auto Resolution, Question-Answering for Search, and Text2Workflow using LLMs. He received ServiceNow's Advanced Technology Group Team of Excellence award for the Text2Workflow product, which is directly analogous to the workflow automation Numos is building for finance. He has co-authored more than twenty publications across the top AI and data conferences. His personal site is mitultiwari.net. For technical candidates, Mitul is the right person to engage directly with a concrete technical take on multi-agent orchestration for financial workflows.