Bengaluru-based OnFinance AI has raised USD 4.2 million in a pre-Series A round led by Surge, the accelerator backed by Peak XV. The funding also drew participation from Groww Founders’ Fund, MarsShot VC, Climber Capital, Indian Angel Network, and Silverneedle Ventures. The company says the capital will help expand its AI engineering team and fuel entry into MENA and US markets.
Why compliance is ripe for AI
Banking and financial services are drowning in regulatory updates. For Indian firms alone, compliance teams must track thousands of pages of circulars and rule changes each year. Globally, the costs are rising: banks spend an estimated 5–10 per cent of operating budgets on compliance, according to industry studies. It is this burden that OnFinance AI wants to ease with domain-specific generative AI.
Its pitch is that general-purpose models are too risky for high-stakes financial environments. Instead, it has trained NeoGPT, a large language model tuned on regulatory and audit data, and packaged it into a platform called ComplianceOS. The system parses rulebooks, generates audit-ready summaries, and flags compliance gaps in workflows — tasks that are still handled manually in many BFSI institutions.
Tech behind the pitch
OnFinance AI says its edge lies in building products tailored for regulation-heavy industries. Its ComplianceOS platform combines workflow monitoring with generative outputs that can be exported for audits or regulatory review. Sitting underneath is NeoGPT, a proprietary large language model trained on regulatory circulars, financial standards, and compliance data.
The company also offers add-on modules, including research assistants for risk teams and agent-based tools that simulate compliance scenarios. Early adopters reportedly include BSE, Kotak Mutual Fund, and HDFC Securities, though these client names have been disclosed in startup announcements rather than independent filings.
By specialising in BFSI compliance, the startup aims to avoid the pitfalls of general-purpose AI, hallucinations, explainability gaps, and audit failures. Whether domain-tuned models can fully deliver on that promise is still an open question.
Market and growth plans
So far, OnFinance AI has focused on India, where regulatory churn is constant and compliance costs cut into margins for banks, brokers, and insurers. The startup says its tools are already being tested by domestic institutions, and the fresh round will allow it to deepen those deployments.
Looking outward, the company has signalled ambitions to enter MENA and US markets. Both regions present different opportunities: the Gulf’s financial hubs are investing heavily in regulatory technology to keep pace with global standards, while US banks face some of the most complex compliance regimes in the world. The funds will also go into R&D and hiring, with an emphasis on AI and engineering talent in Bengaluru and Mumbai.
This trajectory builds on backing from Surge, Peak XV’s accelerator, which has been positioning startups for regional and global expansion.
Risks and trust factors
The promise of AI in compliance comes with real risks. Regulators demand traceability, yet large language models are known to generate outputs without clear provenance. While OnFinance AI says NeoGPT is tuned to minimise hallucinations, independent audits of its effectiveness have not been published.
Trust is also tied to explainability. Financial supervisors need more than black-box answers, and some institutions may be cautious about outsourcing compliance checks to AI. Even in India, where BFSI compliance workloads are heavy, uptake will depend on how comfortable regulators become with machine-generated outputs.
Expansion abroad could heighten these challenges. In the US, for example, financial firms face strict audit requirements, and any errors in AI-driven compliance could carry heavy penalties. Scaling into such markets will require more than funding — it will demand validation and oversight at the highest levels.
What comes next
OnFinance AI’s $4.2 million raise shows how fast generative AI is being pulled into compliance, one of finance’s most conservative domains. By tailoring models to regulation and audit, the company is betting that BFSI institutions will embrace automation if it is narrow, explainable, and regulator-friendly.
The next test is scale. If OnFinance can prove its tools cut costs and reduce compliance risk in India, it could set a precedent for adoption in other emerging markets, and eventually in complex hubs like the US. For now, the startup’s win is less about the size of its round and more about the direction it points: compliance as the next frontier where AI will need to earn trust as much as it delivers speed.
