Senior Data Scientist

Story Terrace Inc.; Bengaluru; On-site; INR 1,200,000 - 1,800,000; Full time

Job summary

Story Terrace Inc. is hiring a Senior Data Scientist in Bengaluru, India to lead the development of predictive AI systems for sensitive financial services including underwriting, fraud monitoring, and risk scoring. The ideal candidate will have significant experience building predictive models, strong skills in classical machine learning, and the ability to navigate the production lifecycle in regulated environments. This role involves building high-performance models and enhancing banking and finance solutions with new AI technologies.

Qualifications

• Significant experience building and shipping predictive models in financial services.
• Strong depth in classical machine learning for tabular data, especially tree-based methods.
• Proven experience with real production use cases such as underwriting and fraud detection.

Responsibilities

• Build and productionize predictive AI models for sensitive financial services workflows.
• Lead model development across structured and tabular datasets using classical ML foundations.
• Own the end-to-end modeling lifecycle including training, validation, and deployment.

Skills

Classical machine learning

Tree-based methods

Model validation

Production monitoring

Risk modeling

Job description

Lexsi Labs builds advanced AI systems for high-stakes enterprise environments. Our work spans aligned and interpretable AI, robust deployment systems, and next-generation model architectures across structured, tabular, and enterprise data.

A major focus area for us is applying AI to sensitive decision systems in banking, insurance, and financial services, where model quality, stability, explainability, and operational discipline matter as much as raw performance.

The Role

We are hiring a Senior Data Scientist to lead the development of predictive AI systems for highly sensitive financial-services use cases such as underwriting, fraud monitoring, anomaly detection, risk scoring, and recommendation systems.

This is a role for deeply experienced practitioners, not generalists and not hobbyists. We are looking for people who have already built, deployed, and managed real predictive systems in production within banking, insurance, lending, or broader financial services.

The core work will involve building high-performance models using classical ML approaches such as GBMs, tree ensembles, calibrated risk models, and structured-data pipelines, while also pushing toward newer approaches using tabular foundation models. This role requires strong judgment on when to use mature methods, when to introduce newer model classes, and how to operate both safely in regulated or high-risk settings.

You will work on systems where errors are costly, model behavior must be defensible, and deployment requires rigor across the full lifecycle: data quality, feature logic, model development, monitoring, drift detection, retraining, synthetic data, evaluation, and production control.

Responsibilities

• Build and productionize predictive AI models for sensitive financial-services workflows, including underwriting, fraud detection, anomaly monitoring, recommendation, and decision-support systems.
• Lead model development across structured and tabular datasets using strong classical ML foundations, including GBMs and other proven approaches for high-signal tabular prediction tasks.
• Evaluate and apply tabular foundation models where they offer real performance or operational advantages over standard approaches.
• Own the end-to-end modeling lifecycle, including problem framing, feature design, training, validation, calibration, error analysis, deployment, monitoring, and retraining strategy.
• Design for real production constraints such as drift, label delay, class imbalance, changing data distributions, noisy supervision, low-default portfolios, sparse events, and regulatory review.
• Build systems for model monitoring, drift detection, continuous training, challenger models, backtesting, and production performance governance.
• Work with synthetic data strategies where appropriate to improve model development, rare-event simulation, privacy protection, and system testing.
• Develop interpretable and defensible modeling approaches suitable for high-stakes environments where model decisions may need to be reviewed by risk, compliance, or business stakeholders.
• Partner with engineering and platform teams to ensure models can be deployed, monitored, audited, and maintained reliably in enterprise environments.
• Contribute to next-generation predictive workflows involving data-science agents and AI-assisted model operations, especially in complex structured-data environments.

What We’re Looking For

• Significant experience building and shipping predictive models in banking, insurance, lending, risk, or financial services.
• Strong depth in classical machine learning for tabular data, especially tree-based methods, ensemble methods, calibration, ranking, and structured prediction workflows.
• Proven experience with real production use cases such as underwriting, fraud detection, anomaly detection, collections, risk modeling, recommendation, or related decision systems.
• Strong command of the full production lifecycle: model validation, monitoring, drift detection, retraining, data quality management, thresholding, and performance tracking.
• Experience working in high-sensitivity or regulated environments where false positives, false negatives, bias, stability, and auditability matter materially.
• Strong product and business understanding, especially the ability to map model behavior to operational and financial outcomes.
• Ability to work with messy enterprise datasets, weak labels, low-frequency events, delayed outcomes, and real-world deployment constraints.
• Genuine curiosity and openness toward tabular foundation models and newer structured-data modeling approaches, while maintaining strong scientific discipline.
• High ownership, sound judgment, and the ability to operate independently on critical systems.

Strong Candidate Signals

• Former Data Science Heads, senior leads, principal-level practitioners, or highly experienced individual contributors from financial-services AI teams.
• Experience deploying models in underwriting, claims, fraud, AML, risk, credit, or recommendation settings at meaningful scale.
• Strong track record of improving production models over time, not just building offline experiments.
• Experience with model drift, challenger frameworks, continuous retraining, synthetic data, and post-deployment performance management.
• Experience working closely with business, risk, and engineering teams in environments where model decisions have real downstream consequences.
• Familiarity with emerging approaches in tabular foundation models and interest in pushing structured-data AI beyond standard feature-engineering pipelines.
• Experience building or supervising complex decision systems that combine predictive models with rules, workflows, human review, and downstream operational controls.

Why This Role Matters

This role sits at the center of how modern predictive AI gets deployed in financial systems.

The work is technically demanding because it spans both worlds: the rigor of proven tabular ML for high-stakes decisions and the frontier opportunity of tabular foundation models as the next layer of capability.

We are looking for people who respect the old machinery, understand why it works, and are excited to build what comes next.