Who Is Kaushal Sheth?

Why It Matters

Sheth’s agentic AI approach could redefine how banks automate trading and risk decisions, offering faster adaptation and stronger compliance audit trails. The shift from static models to learning systems may become a competitive differentiator across the financial services industry.

Key Takeaways

• Sheth co‑founded Sophos Solutions, later acquired by GFT Technologies.
• At GFT, Sheth serves as CTO focusing on agentic AI for finance.
• Otonomii separates memory, reasoning, execution, governance for autonomous financial decisions.
• Continuous learning replaces static prediction, tackling regime shifts and alpha decay.
• Built for traceability, Otonomii logs actions and non‑actions to meet governance.

Pulse Analysis

The financial sector is at a crossroads where artificial intelligence is moving from a supportive analytics tool to a core decision engine. Kaushal Sheth, a veteran technologist with roots in core banking systems, is steering this transition at GFT Technologies. His career, which began with the creation of Sophos Solutions—a platform that powered Latin American banks—now culminates in overseeing the integration of agentic AI across global financial institutions. This evolution reflects a broader industry trend: the need for systems that can not only predict outcomes but also act autonomously within regulatory frameworks.

Agentic AI, the focus of Sheth’s recent work, differs from traditional models by embedding decision‑making capabilities directly into the AI’s architecture. Otonomii, his flagship platform, is built as a modular cognitive system that isolates memory, intelligence, execution, and governance. By continuously ingesting market observations and updating its internal state, Otonomii can detect regime changes, adjust strategies in real time, and maintain a comprehensive audit trail of both actions and inactions. This design addresses persistent pain points such as alpha decay and the difficulty of separating observation from execution, offering a more resilient and transparent AI solution for trading desks and risk management teams.

The implications for banks and asset managers are significant. A learning‑first AI reduces reliance on static, back‑tested models that quickly become obsolete in volatile markets. Moreover, the built‑in governance layer satisfies tightening regulatory demands for explainability and traceability, potentially easing compliance costs. As institutions shift from experimental pilots to production‑grade deployments, Sheth’s architecture could set a new benchmark for AI‑driven finance, prompting competitors to adopt similar autonomous frameworks to stay competitive.