Deriv is a significant case study in the AI agent ecosystem because it represents a large-scale, real-world application of agentic orchestration within a regulated industry. While many companies use AI for simple tasks, Deriv is deploying multi-agent systems for high-stakes fintech operations, including autonomous security auditing in CI/CD pipelines and the transformation of product specs into production code. They specifically categorize this as 'Software 3.0,' moving beyond deterministic and predictive systems into reasoning-based agentic layers.
For those building or using agents, Deriv matters as both a sophisticated consumer of agent technology (integrating tools like Claude Code) and a platform provider. Their public trading API and 'Deriv Bot' no-code builder allow external developers to deploy their own autonomous trading agents. By building an internal 'agentic operating system' to coordinate virtual AI teams, Deriv is pushing the boundaries of how agents can manage complex, coordination-heavy business functions at scale.
Deriv is an online brokerage that has existed since the early days of the commercial internet. Founded in 1999 as Regent Markets Group, the company spent two decades operating as Binary.com before rebranding to its current name in 2020. Based in 21 locations across 16 countries, it manages a trading volume exceeding $700 billion monthly for three million customers. While its core business is the provision of trading platforms for forex and commodities, the company has recently shifted its technical focus toward what it calls Software 3.0.
The shift at Deriv is more than a superficial addition of chatbots. The company is reorganizing its entire technical stack around autonomous intelligence. This stack is split into three layers. Software 1.0 consists of the deterministic code, APIs, and databases that run most fintech systems. Software 2.0 uses machine learning models for predictive requirements, such as risk management. Software 3.0 is where the company focuses on LLM APIs and agent orchestration to handle reasoning and decision-making tasks across its global operations.
One application of this technology is "Amy," an AI support system that turned multi-hour response times into sub-25-second interactions. More complex is the integration of agentic workflows into the software development lifecycle. Deriv uses Claude Code within GitHub Actions to review every pull request for security issues. This system allows developers to query findings and generate code fixes directly within the review flow. The company also builds multi-agent systems designed to transform product specifications into production code. It attempts to automate the path from a technical requirement to a live feature through autonomous virtual teams.
The company describes its vision as the creation of "AI colleagues" rather than simple automation tools. These systems handle millions of customer interactions and make decisions in a high-stakes environment involving real money and strict regulation. This approach places Deriv in an unusual position in the current market. Unlike horizontal AI startups, Deriv is a vertical integrator. It applies agentic patterns to the specific constraints of finance, building an internal infrastructure for autonomous collaboration to solve coordination problems that do not yet have industry-standard solutions or established playbooks.
In the broader market, Deriv sits between traditional retail brokers and newer, tech-focused platforms like eToro. Its primary differentiator is its proprietary Derived Indices—synthetic markets that simulate real-world price movements and are available 24/7. It also maintains an open approach to its trading infrastructure through a public API. This allows third-party developers to build their own trading bots and automated systems. This strategy creates a feedback loop: the company’s internal experiments with agentic systems eventually inform the tools and APIs available to its three million users, who use these resources to automate their own trading strategies.
Programmatic access to trading infrastructure and real-time market data.
No-code automated trading strategy builder.
Deriv is hiring.