Your fractional quant

Stop spending senior research time reviewing AI output.

Send me the idea. I will turn it into a properly specified, tested and challenged research conclusion.

Send a research request See the process

For portfolio managers, investment teams, small funds and family offices.

The problem

AI made generating research cheap. It did not make oversight free.

Someone still has to formalize the hypothesis, check the implementation, challenge assumptions, detect bad methodology, test robustness, account for costs and decide whether the result deserves more time or capital.

Send a hypothesis, strategy, paper, chart, indicator, AI-generated idea, or research question you would otherwise investigate internally. I will return a concise research report with methodology, evidence, robustness checks, risks and a clear verdict.

How it works

Send it. I investigate. You get a verdict.

  1. 01

    Send it

    Forward the hypothesis, strategy, paper, indicator or AI output. Rough ideas are fine.

  2. 02

    I investigate

    I formalize the question, implement the test, challenge the assumptions and run the relevant validation.

  3. 03

    Get the verdict

    Within the agreed turnaround, you receive the evidence, failure modes and my conclusion on whether it deserves further work.

What you can send

Make the first test a real backlog item.

A trading hypothesis, an existing strategy, a paper you want replicated, a market-structure idea, an AI-generated strategy, a factor concept, a portfolio rule, a strategy you suspect is overfit, or a question your team has not had time to investigate.

Click an example to drop it into the request form.

What you get back

A research memo, not a screenshot of a profitable equity curve.

Research specification
Exact hypothesis, assumptions, universe, data and methodology.
Evidence
Baseline results and the evidence that matters.
Validation
OOS testing, sensitivity, costs, robustness and failure analysis where relevant.
Verdict
Reject, investigate further, incubate/shadow, or pursue — with the reasoning behind it.
Research note · RN-24 Anonymized sample

IBS bounce on index futures, lag-1, after costs

Client: family office · Universe: liquid US equity index futures · Horizon: daily

Hypothesis. Close in the lower quintile of the day’s range predicts next-session mean reversion, conditional on a longer-horizon trend filter.

Evidence that matters. The intern backtest’s IS Sharpe is not the result. After costs and lag-1 execution, IS Sharpe compresses sharply. 2018–2020 OOS is positive but concentrated; the 2020 crash window is a wash.

Failure modes. Parameter neighbourhood is unstable below the stated IBS threshold. The trend overlay is doing real work. Fine at index-futures size; not a single-name idea.

Verdict Investigate further — shadow with the trend overlay. Do not allocate on the original specification.

Why me

A researcher who uses AI as leverage, and stays responsible for the conclusion.

I am Vasilis Kotopoulos, a quantitative researcher focused on systematic digital-asset strategies. I have led a six-person quant and data team, built and maintained production trading infrastructure, researched signals and regime models, built robustness-testing and automated strategy-research tooling, and currently run my own research and shadow-incubation workflow.

I use AI heavily myself. That is the point. You are not paying me to manually write code that a model can produce. You are paying for a quant researcher who can use it as leverage while remaining responsible for the research conclusion.

Engagement

One retainer. Bounded capacity. No “unlimited” at launch.

Fractional Quant Research

€500 / month

Pilot retainer · up to four substantive investigations

  • Clearly bounded monthly capacity
  • Written memo with specification, evidence, validation and verdict
  • Turnaround agreed at the start of the engagement
Start with one question

Teams

Custom

Greater throughput, proprietary data, monitoring or deployment

  • Higher investigation volume
  • Integration with internal data and research queue
  • Quoted against actual capacity, not an open-ended retainer
Discuss a team engagement

Send a research request

The lowest-friction test is one item from the backlog.

Send one research question or idea that is currently sitting with the team. The sale, if there is one, comes after you see the quality of the output.

Or email directly: hello@vkotopoulos.gr

Send a research request