Machine Learning Engineer

Sabpaisa; Delhi; On-site; INR 1,500,000 - 2,500,000; Full time

Job summary

Sabpaisa is seeking a Machine Learning Engineer in Delhi to lead the development of innovative ML systems that enhance payment success rates and automate risk assessments. You will manage the entire ML lifecycle, from feature engineering to deployment and monitoring, ensuring real-time detection of fraud. Ideal candidates have 3-5 years of experience developing production models, excellent Python skills, and familiarity with PyTorch or TensorFlow. Engage with product teams for integration of advanced ML models into systems, shaping the future of payments.

Qualifications

• 3-5 years of applied ML experience with models deployed to production.
• Strong background in Python, PyTorch or TensorFlow.
• End-to-end ML skills including feature engineering and deployment.
• Experience with tabular data and classification/ranking problems.
• Understanding of real-time inference and model serving.

Responsibilities

• Build smart payment routing models to optimize transaction success rates.
• Design and deploy real-time fraud detection systems.
• Develop merchant risk scoring models for automated underwriting.
• Build document intelligence pipelines for KYC documents.
• Set up ML infrastructure and monitoring alerts.
• Collaborate with teams to integrate ML models into production.

Skills

Machine Learning

Python

PyTorch

TensorFlow

Data Manipulation (pandas/numpy)

Job description

Overview

You will build ML systems that improve payment success rates, detect fraud in real-time, and automate merchant risk assessment. Every 1% improvement in success rate translates to crores in additional GMV for our merchants.

You'll own the full ML lifecycle: feature engineering, model development, deployment, and monitoring. This is a greenfield role—you'll shape the ML architecture from scratch.

Key Responsibilities

• Build smart payment routing models to optimize transaction success rates across UPI, cards, and net banking rails
• Design and deploy real-time fraud detection systems with sub-100ms latency—device fingerprinting, behavioral analysis, velocity checks
• Develop merchant risk scoring models for automated underwriting using documents, transaction patterns, and external signals
• Build document intelligence pipelines—OCR, classification, and data extraction for KYC documents (PAN, Aadhaar, GST, bank statements)
• Set up ML infrastructure—feature stores, model serving, A/B testing frameworks,monitoringand alerting
• Collaborate with product and engineering teams to integrate ML models into production systems
• Monitor model performance, detect drift, and implement retraining pipelines
• Document model architecture, training procedures, and performance metrics

Requirements

• 3-5 years of applied ML with models deployed to production (not just notebooks/Kaggle)
• Strong Python with PyTorch or TensorFlow, plus pandas/numpy for data manipulation
• End-to-end ML skills: feature engineering, training, evaluation, deployment, monitoring
• Experience with tabular/transactional data and classification/ranking problems
• Understanding of real-time inference—latency budgets, feature stores, model serving

Good to Have

• Fraud detection or risk modeling experience in fintech/payments
• Multi-armed bandits or reinforcement learning for optimization
• Graph neural networks for network-based detection