QSentia · 3 hours ago
Reinforcement Learning Engineer
United States
Part-time
Remote
Mid, Senior Level
1+ years expQSentia is developing a next-generation AI-driven hedge fund platform that integrates reinforcement learning, large language models, and quantitative research. The role involves designing, implementing, and testing RL-based trading and risk models that adapt to market conditions, collaborating closely with AI researchers and portfolio managers.
Financial Services
Hiring Manager
Aditi Dewan
Responsibilities
Develop Advanced RL Agents: Build, train, and evaluate portfolio agents using DDPG, TD3, PPO, SAC, and other actor–critic architectures with risk-aware reward shaping and uncertainty regularization
Model Market Environments: Simulate realistic, event-driven market environments (partially observable MDPs) to test policy robustness under volatility, liquidity stress, and structural breaks
Integrate Alpha & Context Models: Fuse LLM-extracted sentiment and event embeddings (from earnings calls, filings, and news) with numerical alpha factors for multi-modal RL decision-making
Construct Adaptive Risk Overlays: Implement regime-switching and volatility-gated policy layers to manage tail risk, exposure concentration, and position-level drawdowns
Backtest & Benchmark: Design walk-forward, point-in-time backtesting frameworks that incorporate realistic frictions—transaction costs, slippage, and borrow fees—and benchmark agent performance across Sharpe, Sortino, Calmar, and CVaR metrics
Collaborate Cross-Functionally: Work with PMs, ML engineers, and AI researchers to transition prototypes into production-grade portfolio systems deployed on cloud or hybrid GPU environments
Optimize Compute Efficiency: Leverage distributed and GPU-based training to accelerate simulation throughput and policy convergence
Qualification
Required
1–3+ years in quantitative research, algorithmic trading, or reinforcement learning applications in finance
Deep knowledge of RL algorithms (DDPG, TD3, PPO, SAC, A3C), PyTorch/TensorFlow, and policy gradient optimization
Understanding of market microstructure, portfolio optimization, and risk-adjusted performance metrics
Expert in Python (NumPy, Pandas, PyTorch); familiarity with C++ or Rust for performance-critical components is a plus
Experience handling large-scale financial time series, SQL/NoSQL databases, and real-time data APIs (Polygon, Bloomberg, Refinitiv)
Exposure to LLMs (OpenAI, Claude, Gemini, etc.) and text-to-signal modeling for financial event interpretation
Proven ability to eliminate look-ahead bias, apply embargoed cross-validation, and maintain OOS integrity
Preferred
Knowledge of options pricing, implied volatility surfaces, and risk-neutral modeling
Familiarity with distributed training (Ray, Dask, Spark) or simulation orchestration frameworks
Prior research experience at a quant fund, proprietary trading desk, or asset manager
Benefits
Equity participation
Performance-based bonuses
Company
QSentia
Initial back test results: Sharpe: 2.6 Calmer: 5.6 Sortino: 5.5 Max DD: +18%
2-10 employees
Funding
Current Stage
Early StageCompany data provided by crunchbase