Data Scientist – Risk & Fraud Modeling (3–5 Years) - Hand on Coding - Mumbai

Hiring for: An exciting InsurTech startup building an AI-native stack for the Insurance.

Role: Data Scientist – Risk & Fraud Modeling (3–5 Years) - Hand on Coding - Mumbai

Positions: 1

Experience: 3 to 6 years

Location(s): Mumbai

Type: On-site / Permanent

Salary: Up to INR 35 LPA (based on fitment)

About the role

We are looking for a Data Scientist who combines deep BFSI domain expertise with strong machine learning and analytics capabilities. The ideal candidate has extensive experience in fraud, risk, underwriting, and decisioning models, and is excited about applying AI agents and agentic workflows to transform how insurers make decisions. This role will work closely with product, engineering, and customers to build next-generation AI-powered solutions for underwriting, claims, fraud detection, risk assessment, and operational automation.

Key Responsibilities

Data Science & Machine Learning

• Work on end-to-end lifecycle of machine learning and advanced analytics initiatives.

• Design, develop, validate, and deploy predictive and prescriptive models for BFSI and insurance use cases.

• Build scalable data science solutions leveraging structured and unstructured data.

• Model monitoring, performance tracking, explainability, and governance practices.

• Establish best practices for experimentation, feature engineering, model evaluation, and MLOps. Fraud & Risk Analytics

• Design and enhance fraud detection and fraud prevention models.

• Build risk scoring, propensity, anomaly detection, and behavioral analytics models.

• Develop decisioning frameworks for underwriting, claims, collections, and customer risk assessment.

• Partner with business stakeholders to translate risk and fraud strategies into analytical solutions.

• Continuously improve model effectiveness while balancing customer experience and operational efficiency. AI & Agentic Decisioning

• Explore and implement AI agent architectures that assist or automate business decision-making. • Build agentic workflows that combine LLMs, predictive models, rules engines, and enterprise data. • Develop AI-driven copilots and decision support systems for operational teams.

• Evaluate emerging AI technologies and identify practical applications within insurance and BFSI workflows.

• Drive responsible AI adoption with a focus on explainability, governance, and compliance.

Required Qualifications & Experience

• 3 to 5 years of experience in Data Science, Machine Learning, Advanced Analytics, or related fields.

• Strong experience within Banking, Financial Services, Insurance, FinTech, InsurTech, or related domains.

• Proven experience building and deploying machine learning models in production environments.

• Demonstrated experience in fraud analytics, fraud detection, risk modeling, underwriting analytics, claims analytics, or related domains.

• Strong understanding of supervised and unsupervised learning techniques. • Hands-on expertise in Python and common data science libraries (Pandas, Scikit-learn, XGBoost, LightGBM, TensorFlow, PyTorch, etc.). • Experience working with SQL and large-scale data platforms.

• Familiarity with cloud environments and modern MLOps practices.

• Strong communication, customer-facing, and stakeholder management skills.

Preferred Experience

• Experience within the insurance industry.

• Experience with GenAI, LLMs, RAG architectures, AI agents, and agentic workflows.

• Familiarity with decision engines, rules engines, and enterprise workflow platforms.

• Experience leading teams and mentoring data scientists.

• Exposure to model governance, explainability, and regulatory requirements within BFSI. What Success Looks Like

• Delivery of measurable business impact through fraud, risk, underwriting, and decisioning models.

• Successful deployment of AI-powered decision support systems and agentic solutions.

• Strong model governance, scalability, reliability, and production adoption.

• High-quality collaboration across product, engineering, and business teams.