Role Summary
Title: Data Scientist – Applied AI, Personalization & Behavioral Intelligence Location: Gurugram / Remote
Experience: 3+ years
Function: Product & Data
Reporting to: Chief Product Officer (CPO)
Compensation: Competitive; based on experience
Why this role?
You will work at the intersection of behavioral data, ML, and GenAI. Not research for research’s sake. You will ship models that change behavior, drive resolution outcomes, and make help feel human.
You will partner with Product, Design, Engineering, Growth, and Ops to:
– Segment customers and understand behavioral drivers.
– Personalize journeys and communications in-product and via lifecycle tools. – Power AI counsellor experiences that are contextual, safe, and effective.
Key Responsibilities – 1) Data Foundation & Collaboration
– Define the analytics data strategy and partner with Data Engineering to implement it. – Ensure clean pipelines, consistent schemas, and event instrumentation across app, CRM, payments, and bureau data.
– Build and maintain a unified customer 360 / graph to power segmentation and real-time scoring. – Own data quality checks and model-readiness KPIs; you guide, DE builds.
Key Responsibilities – 2) Segmentation & Behavioral Modeling
– Build clustering to classify users by stress level, risk, engagement, and intent. – Ship propensity models (churn/delinquency, settlement likelihood, recovery potential, product affinity).
– Translate insights into journey hypotheses with Product (nudges, offers, education, conversational flows).
– Validate via A/B tests, uplift modeling, and causal inference; close the loop with learnings.
Key Responsibilities – 3) Applied ML & Generative AI in Product
– Train and deploy classification, recommendation, and personalization models (Python, scikit-learn/XGBoost/PyTorch/TensorFlow).
– Use LLMs and embeddings for summarization, sentiment, intent detection, and agent assistance (counsellor scripts, next-best-action).
– Build vector search for content/user-state matching (FAISS/Pinecone or similar). – Work with Engineering to productize via APIs or real-time inference services; own performance thresholds.
– Implement model monitoring (drift, bias, latency, cost) and retraining cadence.
Key Responsibilities – 4) Decision Science & Measurement
– Stand up self-serve dashboards and automated analytics for cohorts, funnels, campaigns, and models.
– Define and own north-star behavioral metrics with stakeholders: activation, engagement, delinquency reduction, settlement rate, recovered value, CSAT/NPS.
– Run deep-dives to explain model impact, inform roadmap, and prioritize features with Product. – Champion ethical AI, privacy, and bias management.
Must-Have Qualifications
– 3-6 years in Data Science / Applied ML with models shipped to production. – Strong unsupervised learning (K-Means/DBSCAN/PCA) and predictive modeling (Logistic Regression, XGBoost/GBMs, etc.).
– Proficiency in Python and SQL; experience with scikit-learn and at least one deep-learning framework (PyTorch/TensorFlow).
– Data warehousing experience (BigQuery/Snowflake/Redshift) and comfort working with DE on pipelines.
– Familiarity with LLM/GenAI frameworks (LangChain/OpenAI/HuggingFace) and vector databases. – Hands-on with experimentation: A/B testing, uplift/causal methods, guardrails, and statistical power.
– Clear, concise communication; turns data into decisions for non-technical partners.
Good-to-Have
– Fintech/BFSI/consumer-tech background with behavioral or credit modeling. – Real-time inference stacks (SageMaker/Vertex/KServe) and feature stores. – Lifecycle orchestration (MoEngage/CleverTap/Braze) and event taxonomies. – Experience with observability for ML (drift, fairness, cost).
– Practical understanding of privacy, consent, and ethical AI in Indian context.
How We Will Measure Success (Illustrative)
– +X percent activation / +Y percent engagement in targeted segments.
– Decrease in delinquency risk in at-risk cohorts; increase in settlement success / recovered value. – Time-to-insight down and model-to-prod cycle time down through better data readiness. – Proven impact via statistically valid experiments and clear decision memos.
What We Offer
– A high-impact role in a mission-driven team.
– Ownership of core ML systems in a user-first, agentic-AI product.
– Freedom to experiment and ship.
– A culture of clarity, curiosity, and empathy.
…