FOUNDER · STRATEVA.AI

Hi, I'm Filip.

MSc Quantitative Finance · Madrid
Quant Researcher · Hedge Fund, Jersey
Risk Consultant · Currently

I'm currently working as a risk consultant, with previous experience as a quant researcher at a hedge fund in Jersey. I've spent too long watching the best financial tools stay out of reach for most people — and I decided to do something about it.

I was born in Romania and moved to Spain as a child. I studied Finance, Banking and Insurance, then completed a Master's in Quantitative Finance — all in Madrid. From there, my career took me through both ends of the industry: front-line retail banking with clients, quantitative research at a hedge fund, and now risk consulting.

That combination gave me an uncomfortable perspective: the models that actually work — GARCH, Black-Litterman, HRP, Monte Carlo — are locked inside Bloomberg terminals and institutional trading desks. The retail investor, the importer with currency risk, the finance student — they all operate with incomplete information, not because they lack ability, but because they lack access.

Strateva was born to change that.

Finance is a black box. That has to change.

Finance has been a taboo subject for decades. The models exist, the data exists, the academic theory is solid and public. But between Markowitz's equation and a retail investor's portfolio there's a gap that nobody has been willing to fully close.

I believe AI can do that — not in the sense of automating decisions, but in making what was once opaque understandable. So a treasurer understands their currency risk. So a retail investor uses the same models as a family office. So risk management stops being exclusive.

We're still far from harnessing the full potential of AI in portfolio management, risk management and screeners. Strateva is my attempt to push in that direction.

Uncompromising rigor

Strateva's models use walk-forward validation with non-overlapping windows and T-1 data to eliminate look-ahead bias. It's not marketing — it's the difference between an honest backtest and one that only works in the past.

No black boxes

Every result shows its methodology, its sources and its limitations. If the model says your worst case is a 5% loss, you can audit exactly how it arrived at that number.

Real access, not symbolic

The free plan includes tools that five years ago only existed on terminals costing €20,000/year. That's intentional.

A fact about me: I'm passionate about Porsche. We can debate their stock performance — but their cars are a different story. Growing up between data curves and asphalt curves isn't so different: in both cases, technical precision makes all the difference.

If you have feedback, an idea, or simply want to discuss whether the Porsche 911 Targa justifies its price — I'm at info@strateva.ai

Frequently Asked Questions
TRUST
Can I trust the model results?
The models are based on peer-reviewed academic literature (Markowitz, Engle, López de Prado). We use walk-forward validation with real data to measure out-of-sample performance. No model predicts the future — but it can quantify risk rigorously and transparently.
What makes Strateva different from a robo-advisor?
A robo-advisor makes decisions for you with opaque models. Strateva gives you the models so you can make your own decisions with full transparency. We don't manage your money — we give you the tools to manage it yourself.
Is there a free trial without a credit card?
Yes. The Free plan requires no credit card and gives permanent access to a real subset of tools. You can evaluate the platform before any payment decision.
Does Strateva provide investment recommendations?
No. Strateva is a platform of analytical and educational tools. I do not offer financial advice, buy/sell recommendations, or portfolio management. Each user is responsible for their own investment decisions. Always consult a regulated financial advisor.
TOOLS
What is the Portfolio Lab?
The Portfolio Lab is our main portfolio optimization tool. It includes 12 optimization models (Equal Weight, Min Variance, Max Sharpe, Risk Parity, HRP, Black-Litterman, CVaR, etc.), backtesting with walk-forward validation, correlation analysis and risk decomposition.
What is the AI Analyst?
The AI Analyst is an artificial intelligence assistant (powered by Claude from Anthropic) specialized in quantitative finance. It can help you interpret results, explain models, answer questions about financial metrics, and guide you through the platform's tools.
What markets do the tools cover?
The tools cover global markets: IBEX35, S&P 500, international stocks, currency pairs (Forex), ETFs, indices, commodities and macroeconomic data. Data availability depends on the provider (FMP covers 70,000+ instruments).
Can I access it from mobile?
Yes. Strateva is responsive and works on mobile devices, tablets and desktop. All tools adapt to screen size, although for complex analyses like Portfolio Lab we recommend a desktop screen.
DATA & PRIVACY
Where does the data come from?
Financial data comes from Financial Modeling Prep (FMP) for fundamental and market data, the Federal Reserve for interbank interest rates, and TradingView for interactive charts. We do not generate or fabricate synthetic data.
How is my data protected?
All communications are made via HTTPS. API keys are stored exclusively on the server (never in the browser). IP addresses are hashed and never stored in plain text. We comply with the GDPR (General Data Protection Regulation). See our Privacy Policy for more details.
What data can I download and what can't I?
Exports (CSV, XLSX, PDF) contain exclusively analytical results calculated by Strateva:

Downloadable: optimization weights, risk metrics (VaR, CVaR, Sharpe), momentum scores, Z-Scores, signals, backtests, equity curves, hedge ratios h*, efficient frontiers, correlations and multi-factor rankings.

Not downloadable: market prices, real-time quotes, financial statements, balance sheets, income statements, trading volumes, insider data or any raw third-party financial data.

Financial data is used as input for our models, but what you download are Strateva's proprietary results and analyses, not the original data.
MODELS
Are the mathematical models reliable?
The models are based on peer-reviewed quantitative finance literature (Markowitz, Black-Litterman, Engle GARCH, López de Prado HRP, etc.). However, no model predicts the future. Past results do not guarantee future returns. All tools are informational and educational and do not constitute financial advice.
What do VaR and CVaR mean?
VaR (Value at Risk) estimates the maximum expected loss at a given confidence level (e.g. 95%). CVaR (Conditional Value at Risk) or Expected Shortfall measures the average loss in the worst 5% of scenarios. CVaR is always equal to or greater than VaR and better captures tail risk.
Privacy Policy
Legal Notice
Cookie Policy
Terms and Conditions of Use