Senior Data Scientist; R&D - AI​/ML Private Credit

Position: Senior Data Scientist (R&D) - AI/ML for Private Credit
Location: Bengaluru

Senior Data Scientist (R&D) - AI/ML for Private Credit

Designation
-Member of Technical Staff

Location:

Bangalore

Experience:

5+ years
Mode of work:
Work from Office

About the Role
We're looking for a Senior Data Scientist to lead R&D initiatives at the intersection of LLMs, information retrieval, and private credit analytics. You'll fine-tune small language models on financial documents, build agentic workflows for multi-step reasoning, and develop production-ready extraction systems that power our AI platform for institutional investors.
This role bridges cutting-edge research with real-world deployment. You'll work closely with Prompt Engineers on hybrid LLM+ML approaches, partner with QA Data on evaluation frameworks, and translate research into detailed specs for our Platform Engineering team. Your models will process thousands of credit agreements daily, requiring both innovation and reliability.
What You'll Do
Model Development & Fine-tuning
Fine-tune Small Language Models on proprietary private credit corpus (credit agreements, indentures, term sheets)
Develop information retrieval systems: semantic search, ranking algorithms, and context-aware retrieval
Build agentic workflows with multi-step reasoning, tool use, reflection, and self-correction capabilities
Train classification models for document type identification, section detection, and entity recognition
Create extraction models: NER for financial entities, relation extraction, structured table parsing
Research & Innovation
Partner with Prompt Engineers on prompt optimization strategies and hybrid LLM+ML approaches
Experiment with latest techniques: RAG architectures, fine-tuning methods (LoRA, QLoRA), model distillation
Present research findings to engineering team and stakeholders monthly (progress, insights, recommendations)
Stay current with academic research and industry developments in NLP, LLMs, and financial ML

Production Readiness & Deployment
Write detailed technical specs for Platform team: model architecture, dependencies, deployment steps, API contracts
Define production readiness criteria: performance benchmarks, edge case handling, failover mechanisms, rollback procedures
Create comprehensive model cards: intended use, limitations, bias analysis, performance metrics, monitoring requirements
Optimize models for production constraints: latency  95%, cost    Evaluation & Quality Assurance
Work with QA Data Teams on model evaluation frameworks and benchmark dataset creation
Build evaluation frameworks with offline metrics (accuracy, precision, recall, F1) and online metrics (user feedback, business impact)
Create benchmark datasets: 1K+ examples per task with expert annotations and inter-annotator agreement analysis
Define task-specific success criteria tied to business outcomes (e.g., covenant extraction accuracy → analyst time savings)
Monitoring & Continuous Improvement
Monitor model performance in production: accuracy drift, latency degradation, error patterns, user feedback loops
Investigate performance degradation: is it data drift, concept drift, or infrastructure issues?
Retrain models quarterly with new data, improved techniques, and expanded coverage of edge cases
Maintain model performance dashboards and alert systems for critical degradation

Required Qualifications
Technical Expertise
5+ years experience in ML/NLP with 2+ years focused on LLMs and transformers
Strong hands-on experience with fine-tuning language models (BERT, RoBERTa, GPT-style models, LLaMA/Mistral)
Expertise in information retrieval: vector databases (Pinecone, Weaviate, Qdrant), embedding models, semantic search
Production ML deployment experience: model serving (Tensor Flow Serving, Torch Serve, ONNX), monitoring, A/B testing
Proficiency in Python ML stack:
PyTorch/Tensor Flow, Hugging Face, Lang Chain, scikit-learn, pandas
Domain & Problem-Solving

Experience with document processing and extraction tasks (OCR pipelines, layout analysis, table extraction)
Ability to translate vague business requirements into concrete ML problem statements
Track record of moving models from research/prototype to production with measurable impact
Strong understanding of evaluation methodology: offline vs online metrics, statistical significance testing
Collaboration & Communication
Experience writing technical documentation for engineering teams (architecture docs, API specs, runbooks)
Ability to present complex technical concepts to non-technical stakeholders
Comfortable working in cross-functional teams with prompt engineers, platform engineers, and QA analysts

Preferred Qualifications
Experience in financial services, credit analysis, or Fin Tech (private credit, leveraged finance, structured products)
Familiarity with agentic frameworks:
Lang Graph, AutoGPT, ReAct patterns, tool-calling workflows
Knowledge of model compression techniques: quantization, pruning, knowledge distillation

Experience with MLOps tools: MLflow, Weights & Biases, DVC, feature stores
Understanding of financial document structures: credit…