RiskCurve

I was glad to share my insights on the development of AI in South Korea in a recent report for the Kazakhstan national TV channel Khabar 24 🇰🇿.

We discussed our startup’s journey and touched upon the evolving AI regulatory landscape across different global regions. It’s encouraging to see Kazakhstan taking the future of AI so seriously at the highest national level.

https://youtu.be/v_Q_lrjuE8M?si=j3L1RBKatDPHHVEr

Moving forward! 🚀

I’ve just published a short research note on Day-of-Year behavior in EUR/USD.

Instead of revisiting the usual monthly seasonality stories, the article looks at the market one calendar day at a time, using daily data from 2005–2025. The main result is simple: most days are pure noise. A very small number behave slightly differently, often around liquidity shifts and institutional timing.

The focus is on interpretation and structure, not trading rules.
Detailed tables and diagnostics are documented separately for those who want to dig deeper.

https://riskcurve.ai/blog/day-of-year-seasonality-in-eur-usd

Blog RiskCurve provides market regime detection, volatility analysis, and probabilistic trading context using deep learning to help traders decide when not to trade.

We deployed our first benchmark LSTM model yesterday, covering EUR/USD, GBP/USD, and USD/JPY.

It’s the first real checkpoint on the road toward our broader goal — turning research into something reliable, measurable, and honest about its limits.
Plenty of work ahead. This is just the beginning.

I just published my first research article on Medium.
It’s a statistically disciplined look at Month-of-Year seasonality, using EUR/USD daily data from 2005–2025 — and it reaches a conclusion that challenges a lot of common trading narratives.

This is the first piece in a research series focused on what actually survives testing in markets.

Busting Monthly Seasonal Patterns in Trading EUR/USD What Happens When You Actually Test Them

Most trading strategies don’t really “stop working” overnight.
Usually the market conditions change — volatility shifts, regimes change, risk quietly increases.

We’re building RiskCurve around that idea:
focusing on market context instead of signals or predictions.

Things like:

volatility regimes

market regimes

risk environment

No buy/sell calls — just awareness.

How do you usually decide when not to trade?
Do you change strategy behavior when volatility or conditions shift?

Website: https://riskcurve.ai

Curious to hear how others approach this.

The idea for this platform had been in my head for a long time, but back then I lacked both the knowledge and a clear starting point. That’s why I decided to dive into Data Science and ML.

15 months flew by faster than I ever expected. Today, I have a team, a clear vision, and the knowledge needed to finally turn this idea into reality.

We’re building riskcurve.ai — a platform that uses data and machine learning to help better understand market conditions, risks, and probabilities.

This is just the first step on a long journey, but every great path starts with the first step.

Идея этой платформы жила у меня в голове очень давно, но тогда не хватало ни знаний, ни понимания, с чего начать. Поэтому я принял решение пойти в Data Science и ML.

15 месяцев пролетели незаметно. Сегодня у меня есть команда, чёткое видение и знания, чтобы воплотить идею в жизнь.

Мы создаём riskcurve.ai — платформу, которая с помощью данных и машинного обучения помогает лучше понимать рыночные условия, риски и вероятности.

Это только первый шаг на длинном пути, но любой большой путь начинается с первого шага.