Your resume and cover letter, tailored to exactly what they’re hiring for.
Get past ATS filters
Job summary
A leading data science firm is seeking an Artificial Intelligence & Data Talent Partner to design advanced ML models and enhance transparency through explainability techniques. The role involves collaborating with product and engineering teams to deploy solutions while requiring strong expertise in applied machine learning, quantitative finance, and proficiency in Python and SQL. This full-time, remote position expects candidates to have a background in financial data and blockchain-related analytics.
Qualifications
5+ years of experience in applied ML, quantitative finance, or credit-risk modeling.
Strong Python and SQL skills.
Experience with ML frameworks like PyTorch, TensorFlow, scikit-learn, or XGBoost.
Hands-on experience with financial data.
Familiarity with blockchain and DeFi data.
Experience deploying ML models into production (APIs, orchestration, or streaming systems).
Design and build hybrid ML models to extract insights from unstructured data.
Add explainability to models using techniques like SHAP and LIME.
Build scalable data pipelines and integrate various data sources.
Develop anomaly detection tools and evaluation frameworks.
Work closely with Product and Engineering to ship models.
Backtesting model outputs against historical NAVs, trades, and redemptions.
Collaborating with Product and Engineering to ship models via APIs, SDKs, and dashboards for traders and risk teams.
Partnering with CTO on long-term governance, transparency, and AI ethics.
Skills
Applied machine learning
Quantitative finance
Credit-risk modeling
Python
SQL
Machine learning frameworks
Production ML deployment
Data pipelines
API/SDK delivery
NLP/GenAI tools
Tools
PyTorch
TensorFlow
scikit-learn
XGBoost
LangChain
HuggingFace
Job description
Artificial Intelligence & Data Talent Partner at Discovered MENA 🤖
Remote Role
What you’ll be working on:
Designing and building hybrid ML models that combine supervised learning, time-series forecasting, and NLP to extract insights from unstructured data like PDFs, fund memos, and regulatory filings
Adding explainability to models using techniques like SHAP, LIME, and feature attribution so outputs are transparent and human-readable
Building scalable data pipelines across off-chain fundamentals, on-chain activity, and macro benchmarks
Integrating data from sources like FRED, PitchBook LCD, Securitize, Centrifuge, Maple, and TrueFi, with strong data lineage and freshness guarantees
Developing anomaly detection and reconciliation tools across issuer, administrator, and blockchain datasets
Creating evaluation frameworks to measure accuracy, confidence intervals, latency, and data quality
Backtesting model outputs against historical NAVs, secondary-market trades, and redemptions