Machine Learning Engineer- Model Validation

IO Data Labs is hiring for a Machine Learning Engineer - Model Validation (Trading & Time Series). we are Canada based Organization seeking for a candidate with strong technical background. Apply Now

Note: Model Validation (Trading & Time Series) is mandatory.

NO Freshers, No interns, only experienced candidate is required.

Job Title: Machine Learning Engineer

Required Experience: 4-6 years

Location: Remote

Timing: 6:30pm-2:30am

We work with founders, investors, and early-stage startups to independently validate AI models before they go into fundraising or customer-facing deployments. Our goal is simple:

provide objective, third-party verification that an AI model performs exactly as claimed.

We are now looking for a Machine Learning Engineer with deep experience in time-series modeling and trading systems to help us evaluate proprietary trading models. The results you produce will directly support VC due-diligence and investment decisions.

If you enjoy building evaluation pipelines, breaking down model claims, running real backtests, and exposing whether a model works in reality (not just theory), this is the perfect opportunity for you.

About the Role

You will work closely with our internal team to:

1. Deploy ML Models

• Set up the models in a reproducible cloud or on-prem environment
• Containerize using Docker and expose clean inference interfaces
• Recreate the startup's exact runtime and dependencies for accurate comparison

2. Build Automated Validation Pipelines

• Create workflows for data ingestion, preprocessing, rolling-window inference, and metric generation
• Automate everything using Python, MLflow/W&B, or similar tooling
• Maintain full traceability of datasets, model versions, and configurations

3. Run Backtests Across Market Regimes

Back testing is the core of this project. You will:

• Use historical market data (equities, crypto, or multi-asset depending on the model)
• Test the strategy across multiple conditions: bull/bear/sideways markets, high-volatility periods, black swan events
• Apply realistic assumptions (slippage, fees, spreads, liquidity constraints)
• Evaluate profitability metrics:
• Sharpe, Sortino
• CAGR, Profit Factor
• Win Rate, RR ratio
• Evaluate risk metrics:
• Max drawdown
• Time-under-water
• Tail-risk exposure

4. Stress-Test and Validate Claims

• Detect any sources of bias:
• lookahead bias
• survivorship bias
• data leakage
• overfitting to specific market regimes
• Reproduce performance figures claimed by the model developer
• Run out-of-sample tests and walk-forward validation
• Analyze stability and robustness under shifting conditions

5. Produce VC-Ready Reports

Your work will be used externally, so clarity and neutrality matter. You will:

• Prepare detailed, audit-ready reports for investors
• Present findings in a clear, reproducible, data-driven manner
• Highlight strengths, weaknesses, risks, and assumptions
• Support Q&A during due diligence if required

This is not just engineering work — it's technical truth-seeking.

What We're Looking For

Must-Have

• Strong Python skills (NumPy, Pandas, Scikit-learn)
• Experience with ML frameworks (PyTorch/TensorFlow)
• Prior work in algorithmic trading, quant research, or ML-based time-series forecasting
• Hands-on experience running backtests and analyzing trading strategies
• Solid understanding of OHLCV data, market structure, indicators, execution, and slippage
• Ability to set up reproducible ML environments (Docker, cloud deployment)
• Ability to write clear, structured technical reports

Nice to Have

• Experience with Backtrader, Zipline, vectorbt, or custom backtesting engines
• Familiarity with MLflow, Weights & Biases, ClearML
• Understanding of risk analysis, portfolio theory, and performance attribution
• Prior involvement in model audits, compliance checks, or due-diligence work

Why This Project Is Unique

• You act as an independent evaluator, not a model builder
• Your output directly influences VC investment decisions
• High autonomy: you design the pipeline and testing process
• Deep, meaningful work — uncovering how ML models behave in real markets
• Flexible, remote, high-impact engagement

This is ideal for engineers who love rigorous analysis, transparent evaluation, and working on real-world trading models rather than toy datasets.

Job Types: Full-time, Permanent

Pay: ₹120, ₹150,000.00 per year

Work Location: Remote