Autonomous AI agents for financial markets.
AI agents that monitor, analyse and evaluate financial market data, so traders and financial teams can make better-informed decisions.
What it is
What is an autonomous AI agent in finance?
An autonomous AI agent is software that watches approved information sources, reasons about what it finds against goals you define, and delivers the result, such as an alert, an evaluation or a research report, without someone repeating the same steps by hand. Fellow Data's agents analyse and inform; they never place trades or give investment advice.
Financial markets generate enormous volumes of information. Fellow Data's agents are designed to process it continuously and take on the monitoring and research work your team repeats every day, so your traders can spend their time on their own decisions.
Every agent runs inside a written policy. Our AI governance policy explains the principles behind it.
How agents work
Observe, analyse, evaluate and report.
Six capabilities, each limited by the agent's policy. Trading decisions stay with your traders.
01
Observe
Continuously monitor approved information sources.
02
Interpret
Combine structured and unstructured information.
03
Reason
Evaluate information against predefined objectives and rules.
04
Evaluate
Assess scenarios, risk and possible market impact.
05
Inform
Deliver alerts, briefings and research to your team.
06
Report
Keep records of the analysis and outputs.
Agents
Agents for the work your team repeats.
Each agent has a defined role, its own data permissions and its own policy.
Market intelligence agent
Monitors approved market information, including structured data from markets data analytics, and tells you what changed and why it may matter.
- Economic events
- Market news
- Price movements
- Volatility and correlations
- Macroeconomic indicators
Risk monitoring agent
Continuously evaluates risk against your limits and flags conditions that need attention.
- Exposure and positions
- Risk limits
- Unusual activity
- Concentration
- Operational conditions
Trading research agent
Assists traders with research and keeps their analysis documented.
- Market research
- Technical and fundamental analysis
- Scenario analysis
- Historical comparisons
- Trade documentation
Portfolio monitoring agent
Tracks portfolio risk between reviews, so changes don't wait for the next meeting.
- Portfolio exposure
- Asset allocation
- Drawdown
- Volatility and correlations
- Concentration
Compliance monitoring agent
Potential workflows that support compliance teams with monitoring and documentation.
- Rule monitoring
- Exception identification
- Surveillance support
- Documentation
- Audit preparation
AI agents operate within configurable permissions and human-defined workflows. They provide analysis and information to support your team's own decisions; they do not trade or give investment advice.
Agent console
See an agent at work.
Market Sentinel detects a volatility event, gathers context, evaluates the risk and sends a briefing to the trading desk. Every step is written to the audit log.
Our principle
Autonomous does not mean uncontrolled.
Fellow Data's agents work autonomously on analysis, never on trading. Choose a level to see how much analysis an agent may do on its own.
Level 3: Evaluate
The agent evaluates scenarios and risk, and explains what it found.
- Allowed
- Reviewed by your team
- Within configured limits
- Not available
- Read approved dataAllowed
- Generate analysisAllowed
- Send alertsAllowed
- Evaluate scenarios and riskAllowed
- Prepare research reportsNot available
- Run multi-step researchNot available
- Place or prepare tradesNot available
- Give investment adviceNot available
- Audit loggingAllowed
Fellow Data's AI agents analyse and evaluate information only. They do not place, prepare or execute trades, and they do not give investment advice. Trading decisions always stay with your traders.
Policy engine
Permissions you can read at a glance.
Each agent runs under a written policy: which data it may read, which tools it may use, what your team reviews and what it may never do. Agents have read-only access: they cannot trade, move money or change limits.
Safety architecture
Every output passes through a policy layer.
Agents get read-only access to the data their policy allows, and no access to place trades or move money. Every piece of work takes this path.
AI agent
Receives an objective and the data it is authorised to use.
Tools
Only the tools registered for this agent are available to it.
Policy engine
Checks every task and output against the agent's written policy.
Permissions
Confirms the agent may read this data and use this tool.
Scope limits
Keeps the analysis within the scope you configure.
Review
Sends outputs that need checking to a member of your team.
Output
Delivers the result, such as an alert, an evaluation or a research report.
Audit
Records the event, the decision, the policy applied and the result.
AI-generated outputs may be incomplete or inaccurate and should be reviewed according to the intended use and applicable controls.
Questions
Questions about AI agents.
Can a Fellow Data agent place trades?
No. Fellow Data's agents monitor, analyse and evaluate market data and deliver the results to your team. They do not place, prepare or execute trades, and they do not give investment advice. Every trading decision is made by your traders.
Do AI agents replace traders or risk managers?
No. Agents take on continuous monitoring and repeatable research so that people can focus on decisions. Your traders make every trading decision and stay accountable for it.
Are AI agent outputs always correct?
No. AI systems can produce incomplete or inaccurate outputs, and they should not be treated as guaranteed predictions of market behaviour. Outputs should be reviewed according to their intended use and your controls.
What data can an agent access?
Only the sources its policy authorises, such as market data, news or portfolio data. Each source is granted explicitly in the agent's policy, and every access can be recorded in the audit log.
Keep exploring.
Risk management technology
Route client orders to A-book or B-book by rules, watch net exposure per symbol, profile toxic clients and enforce risk limits with Fellow Data's risk software.
Explore risk management technologyMarkets data analytics
Analyse COT reports, open interest, SEC filings, forex, equities, commodities and macroeconomic data in one environment with Fellow Data markets data analytics.
Explore markets data analyticsAI governance
Fellow Data's responsible AI principles and agent control framework: human accountability, oversight, permissions, risk limits, logging and transparency.
Explore AI governance
Put AI agents to work, under your rules.
Tell us which monitoring, analysis or research work takes your team the most time.