Bengaluru-based generative AI startup OnFinance AI is moving in on the Middle East’s banking and regulatory sector, founder and CEO Anuj Srivastava told EnterpriseAM MENA <> India in an exclusive interview. The firm will accelerate the rollout of its GCC-specific AI model and explore expansion into markets where agreement sizes are 2x to 3x larger than in India.

**EnterpriseAM Mena<>India is our latest briefing tracking one of the world’s most dynamic trade, investment, and cultural corridors, every Monday, Wednesday, and Friday. You can subscribe to EnterpriseAM MENA <> India by tapping here.

Marquee clients: Founded in 2023, the startup currently serves more than 25 Indian banking, financial services, and ins. institutions, including the BSE, HDFC Securities, Kotak Mutual Fund, Nippon Mutual Fund, and Aditya Birla Capital Digital, among other heavyweight mutual funds.

A GPT for the Gulf region in the works: Central to OnFinance’s Middle East expansion strategy is the NeoGPT-GCC, a fine-tuned large language model trained on regulatory data from over 40 authorities, including the Saudi Central Bank (Sama), the Central Bank of Kuwait, and the Central Bank of the UAE. The model is also trained with data from multi-sector regulators including Saudi Arabia’s Capital Market Authority, the UAE’s Securities and Commodities Authority, and the Financial Services Regulatory Authority in Abu Dhabi.

How does it work? The model parses Arabic circulars, legacy PDFs, and digital and scanned regulatory documents to convert them into “plain English action tables” for compliance and audit teams, said Srivastava. OnFinance claims a translation accuracy rate of 98%. Built on a modular, rules-driven architecture, the system supports local identity formats, sanctions lists, and region-specific workflows across banks, insurers, and capital-markets entities.

Strategic entry: With the NeoGPT GCC already live, a formal market rollout and soft launch of the LLM is slated for late December, Srivastava told us. The firm plans to hire two Gulf-based team members to support on-premise implementations.

OnFinance AI secured USD 4.2 mn in a pre-series A round in September led by Peak XV’s Surge, one of India’s largest VC funds, along with participation from Groww Founders’ Fund, MarsShot VC, Climber Capital, and existing investors Indian Angel Network and Silverneedle Ventures. The money will fund OnFinance’s expansion into the Middle East, Southeast Asia, and US markets to advance its BFSI-focused LLM research.

Funding: OnFinance is also seeking funding from several Gulf-based private offices, including the Seed Group, a company under the private office of UAE’s Saeed bin Ahmed Al Maktoum. The company is also interested in working with consultancy firm Protiviti to approach GCC-based investors.

The company revamped its go-to-market strategy after pitching early proofs-of-concept to First Abu Dhabi Bank and Arab National Bank last year. The plans stalled due to product immaturity at the time, but Srivastava says the platform is now production-ready for the region’s finance industry.

Data sovereignty and security: With GCC institutions insisting on data residency, OnFinance is going ahead with full on-premise deployments rather than going for the cloud. The company says it has built enterprise-grade governance tools including data-lineage tracking, model-drift monitoring, policy-enforcement guardrails, and role-based access. “Banks in the Middle East are extremely bullish on GenAI, but uncompromising on data security,” Srivastava said, noting that GCC firms prefer vetted open-source LLMs over closed-source AI stacks.

OTHER STARTUP NEWS-

US–based robotics startup Buildroid AI is rolling out its block-laying construction humanoid in the UAE after raising USD 2 mn in a pre-seed funding round led by venture capital investor Tim Draper, who has backed Tesla, Robinhood, and Skype, according to a press release (pdf). The firm plans to use the capital to scale its pilot programs, improve the simulation and autonomous functions of its humanoid crew, and push ahead with the commercial deployment of its block-laying machines.

Buildroid AI has already tested its robots on active construction sites in the UAE — including one of Alec Construction’s — and aims to launch its first commercial AI robotics teams with contractors in 2Q 2026.

Why the UAE? Founder Slava Solonitsyn touted the UAE’s appealing compliance landscape and strong demand from contractors for skilled workers, along with a gap in talent supply, in an emailed statement to EnterpriseAM UAE.

The block-laying humanoid targets a market gap in AI-driven construction, where existing single-task construction robots tend to struggle in complex environments and need human support, limiting productivity and returns, according to the release. The firm plans to integrate its machines with Autonomous Mobile Robots that can transport concrete blocks directly from pallets to the humanoids — automating up to 80% of masonry work. The startup sees the role of construction workers evolving in five to ten years to one that is more supervisory and limited to more nuanced, professional tasks like final-touch operations, Solonitsyn told EnterpriseAM UAE.

More plans in the pipeline: The company plans on expanding its robot applications to robotic operators, vendors, and contractors, employing an ERP simulation-driven planning system to optimize efficiency.

Tags: