Your market brainCapture the way you see markets, not someone else's indicator.
Personal strategy research agents
Turn your trading brain into a research agent.
BayesFlow helps traders encode the way they see markets, search for similar historical behaviours, and test whether those patterns are statistically repeatable.
Private technical prototype
Personal research agents
No live execution
BayesFlow is currently a private research and analytics prototype. It does not provide financial advice, guarantee trading performance, or offer public live execution.
Find what mattersSearch historical behaviour using your logic and context.
Test with evidenceTurn candidate trading patterns into simulations and comparable metrics.
Improve over timeAgentic review flags weaknesses and surfaces new ideas.
Stay in controlYou remain central while BayesFlow scales the research work.
Intuition is powerful. Evidence is scalable. BayesFlow connects them.
BayesFlow turns trader intuition into a repeatable research ecosystem.
The research bottleneck
Less digging. More creative strategy work.
As a professional proprietary trader, one of my biggest frustrations is the time spent manually confirming whether a pattern repeats.
What if you could put your trading mind into a research worker that performs the digging and verification for you?
BayesFlow does the digging so traders can focus on discovering, combining, and challenging strategy ideas.
Early research proof
Agent-found patterns, ready to test.
These related intraday behaviours were surfaced during an early agent-assisted research pass.
Find the behaviour. Preserve the evidence. Test whether it repeats.
Example agent outputPrototype · unvalidated
PF-001 · continuation structure · medium confidence
- Input
- NDX.INDX · 5m · 79 candles
- Found
- 1 repeatable candidate family
- Observed
- 4 occurrences · 4 positive reactions
- Trace
- Payload, retrieval, rejected candidates, limitations
Agent candidate 01Unvalidated example
How it works
From intuition to evidence in one research loop.
Teach the agent how you see the market. Let evidence challenge the idea.
Capture
Share annotated charts, trading notes, setup rules, and market context.
Encode
Turn that trading research into your personal research profile.
Search
Find related historical behaviours.
Test
Run structured simulations and compare outcomes.
Learn
Use false positives, valid patterns, and missed setups to improve.
Founder story
Built from obsession. Designed for evidence.
“Being a solo trader is not easy. The largest funds spend billions on research, data, infrastructure, and analytics every year, while independent traders are often left with screenshots, journals, and intuition. I started BayesFlow because I wanted to give serious traders something closer to their own research team — an agentic research environment that can investigate specific market patterns, search for similar behaviours, test their feasibility, and help turn trading ideas into evidence.”
BayesFlow began with a simple obsession: understanding why markets move the way they do.
Hi, I'm Max. After years of studying intraday price behaviour as a speculator, and now trading professionally within a proprietary environment, I became focused on a problem at the centre of discretionary trading.
Skilled traders often see patterns before they can explain or prove them. The problem is turning those observations into structured hypotheses that can be searched, tested, measured, and improved without spending countless hours reviewing markets manually.
BayesFlow was created to change that.
BayesFlow turns a trader's pattern-recognition process into a personal research agent.
The first BayesFlow agent is being built around my own intraday pattern-recognition process. Once the harness is proven, the same framework can support other traders, educators, and research teams.
Roadmap
Prove the harness. Then scale the agents.
- In progressFounder agent + testing foundation
Prove the research loop around Max's own process.
- NextReusable agent harness
Turn annotated charts, trading notes, setup rules, and market context into personal research profiles.
- PlannedControlled private beta
Test with selected traders, educators, and research teams.
- LaterTeam workflows + governed execution research
Only after evidence, licensing, compliance, and risk controls.
Funding and partnership
Why support matters now.
BayesFlow has moved from idea to technical prototype. Support now helps turn the founder-built research loop into a secure, scalable private beta.
Discuss funding / partnershipCloud infrastructureSimulation workloadsAgent evaluationSecure private beta
Private research access
See the technical prototype.
The BayesFlow dashboard is access-controlled because it contains proprietary strategy research workflows. Reviewer access can be provided on request through a read-only demo environment using sample data and no execution features.
max_h@bayesflow.xyz