AI agents are everywhere. Reliable, enterprise‑grade AI agents for high‑stakes finance? That’s a much shorter list.

Samaya AI, a fast‑rising AI platform focused exclusively on financial services, has introduced what it calls an Agent Control Plane (ACP)—a purpose‑built architecture designed to create, run, and govern next‑generation AI agents capable of reasoning in real time with deep domain context. Alongside the launch, the company announced new investment from NVentures, NVIDIA’s venture arm, and Databricks Ventures.

The combination of a new technical blueprint and heavyweight backers signals Samaya’s ambition: to become the institutional intelligence layer for financial services.

Why Generic AI Agents Fall Short in Finance

In consumer and general enterprise use, AI agents typically operate within narrow workflows—summarizing documents, drafting emails, or executing basic tool calls. Finance is different.

Investment professionals must:

• Synthesize structured and unstructured data
• Reason across millions of datapoints
• Pull from internal research, market feeds, filings, and proprietary models
• Maintain audit trails and explainability
• Deliver conclusions that can withstand regulatory scrutiny

According to Samaya CEO and Founder Maithra Raghu, generic AI agents “break down under this complexity—hitting context limits, compounding errors, and degrading as tools scale.”

Samaya’s ACP is designed to address that failure mode head‑on.

Inside the Agent Control Plane (ACP)

At its core, ACP is an architectural layer that allows financial institutions to design, customize, and govern AI agents tailored to specific workflows—whether that’s earnings analysis, macro scenario modeling, or portfolio risk evaluation.

Unlike off‑the‑shelf AI copilots, ACP allows users to:

• Create agents via natural language instructions
• Iteratively refine behavior in real time
• Control how agents plan, execute, and analyze tasks
• Maintain transparency into how conclusions are generated

The architecture integrates several proprietary modules developed by Samaya:

• An integrated Planner
• A Long Horizon Executor
• A Memory module
• A Reasoning module
• Context Management and Optimized Tools modules

The inclusion of long‑horizon execution and memory is especially notable. Many AI systems struggle with extended, multi‑step reasoning over large datasets. In financial modeling or enterprise research workflows, tasks often span multiple analytical layers and data sources. Maintaining coherence across those layers is a non‑trivial engineering problem.

By tightly integrating planning, reasoning, execution, and memory, ACP attempts to reduce context fragmentation and error accumulation—two common failure points in multi‑agent systems.

Built for Auditability and Institutional Control

In finance, autonomy without governance is a non‑starter.

Samaya emphasizes that ACP is designed with auditability and transparency in mind. Agents must not only generate conclusions but also make it clear how those conclusions were reached. That requirement reflects increasing regulatory focus on model risk management and explainability, particularly in banking and asset management. AI outputs that influence trading strategies, credit decisions, or investment recommendations must be traceable.

Samaya’s ACP positions governance not as an afterthought but as a built‑in control plane—hence the name.

Already at Enterprise Scale

This isn’t a concept‑stage platform.

Samaya reports that it is already in production with more than 10,000 professionals at one of the world’s largest banks. Financial services clients are deploying ACP at enterprise scale for use cases including:

• Comprehensive earnings analysis
• Economy‑wide scenario modeling
• Investment research synthesis
• Cross‑asset data analysis

Crucially, ACP integrates internal enterprise data with external content sources and analytical tools. That unified context is what enables agents to produce outputs aligned with an institution’s proprietary methodologies, risk tolerances, and research frameworks.

Rather than acting as a generic AI layer, ACP is meant to operate within—and adapt to—each firm’s institutional context.

Strategic Backing From NVIDIA and Databricks

The new funding from NVentures and Databricks Ventures adds strategic weight.

NVIDIA’s involvement suggests alignment with high‑performance AI infrastructure and model optimization—areas increasingly critical as financial AI agents reason across massive datasets. Databricks Ventures’ participation signals synergy with data engineering and analytics pipelines, particularly in enterprises that rely on unified data lakehouse architectures.

Together, the backing reinforces Samaya’s positioning as infrastructure, not just application software.

The funding will support continued development of ACP and expansion of customizable AI agents for financial workflows.

A Crowded Field—But a Clear Niche

Samaya is entering a rapidly intensifying race.

Major banks are building internal AI copilots. Big tech firms are pitching foundation models fine‑tuned for finance. Fintech startups are embedding AI across research, trading, and compliance stacks.

What differentiates Samaya is its claim to domain specificity and architectural depth. Rather than offering a generalized LLM wrapper, it is presenting ACP as a blueprint for how AI agents should be structured in highly regulated, data‑dense environments.

If that blueprint proves resilient under enterprise stress tests, Samaya could carve out a defensible niche as the control and orchestration layer for financial AI agents.

If not, it risks being absorbed into broader AI platforms that are rapidly expanding their financial capabilities.

The Bigger Shift: From Copilots to Institutional Agents

The launch of ACP underscores a broader transition in financial AI.

The first wave focused on productivity: summarizing research, drafting memos, answering knowledge queries.

The next wave is about decision intelligence—systems that reason across global information flows and contribute directly to investment and risk decisions.

That shift demands more than better language models. It requires orchestration, memory, explainability, and governance at scale.

Samaya AI is betting that the future of financial services AI isn’t just smarter models—it’s smarter control planes.

Get in touch with our fintech expert