Rogo launches Felix, an autonomous agent for bank and investment firm workflows

Rogo has launched Felix, an autonomous AI agent the company says is now deployed across what it describes as many of the world's largest banks and investment firms. The product extends Rogo's existing position as a real-time assistant for financial professionals into the agentic tier: software that initiates and completes multi-step workflows on its own, rather than waiting for a prompt.

The pitch is that Felix runs inside the tools analysts already use, email, messaging, spreadsheets, and data terminals, and handles research, analysis, document preparation, and deal-process management asynchronously. Rogo frames it as a colleague that anticipates the next ask and works across every deal and portfolio simultaneously, with enterprise-grade security and data governance.

The announcement also names Kevin Buehler as Rogo's Chief Innovation Officer, though the public excerpt of the post cuts off mid-bio. Buehler is a recognizable name from McKinsey's risk and AI practice, but the source as published does not detail his prior role or remit at Rogo, so we are not asserting either here.

Rogo's framing of Felix as an agent that produces "real deliverables like decks and models" and embeds inside institutional systems is the standard 2025 vendor pitch for finance-specific AI: move past chat, ship outputs, sit inside the workflow rather than next to it. The company also explicitly references the economic argument, leverage that shows up in the P&L, which is the language buyers at heads-of-research and COO levels actually want to hear.

What the announcement does and doesn't say

The post is a positioning document, not a disclosure. Rogo does not name any specific bank or fund using Felix, does not quantify deployment count, and does not specify which data terminals are integrated (Bloomberg, FactSet, S&P CapIQ, and Refinitiv all sit in this category and have very different integration postures toward third-party AI). It also does not address how Felix handles the compliance constraints that make autonomous agents harder to deploy in regulated finance than in, say, software engineering: audit trails, MNPI segregation, model output review, and the question of who signs off when an agent generates an IC-bound deliverable.

Those are the questions a buyer-side evaluation will turn on. The product narrative is well-constructed; the proof points are not yet public.

What to ask next

For analysts evaluating the category, the useful questions are specific. Which terminals does Felix actually read from and write to? What is the human-in-the-loop checkpoint structure for outputs that touch a live deal? How does Rogo handle MNPI walls when the same agent has read access across multiple deal teams? What is the pricing model, per-seat, per-workflow, or consumption, and how does that interact with the leverage claim?

Those answers, not the launch post, will determine whether Felix is the colleague Rogo describes or another assistant that ends up parked next to the existing stack.