- Staff Data Scientists bridge research ambition and production reality.

- You are the person who takes a promising model architecture and gets it into production reliably, measurably, monitored, and iterable.

- You own the full ML lifecycle for a domain: problem formulation, data pipeline, feature engineering, model selection, A/B design, deployment, and monitoring.

- You own models that run 24x7 in the booking path production breakages have real user impact

- You work across multiple platforms simultaneously your features feed models on Flights AND Hotels AND Bus

- You mentor Senior Data Scientists; your code review raises the whole team's engineering quality

- You are the accountability layer between research exploration and production reliability

Core Responsibilities :

End-to-End ML Ownership :

- Own the full ML delivery cycle: problem definition feature design model training A/B test production deployment monitoring iteration

- Design production-grade feature pipelines: real-time streaming features (Kafka Streams/Flink) + batch features (Spark) with data quality checks and lineage tracking

- Build model evaluation frameworks: offline metrics (NDCG, MAP, AUC, RMSE) + online metrics (CTR, conversion, revenue per session) with alignment analysis

- Deploy models to low-latency serving infrastructure (TorchServe, Triton) with shadow mode testing, canary releases, and automated rollback triggers

AI Domain Delivery (Cross-Platform) :

Ranking: implement and iterate on L2R models for search results across Flights, Hotels, Bus, Train including position bias correction and diversity constraints

Recommendation: deploy collaborative filtering and session-based recommendation models; own the online/offline metric alignment

Price ML: build fare prediction and price alert systems; calibrate demand forecasting models with seasonal adjustments for Indian travel patterns

NLP pipelines: deploy query understanding classifiers, intent detection models, and entity extraction for search across all verticals

LLM integration: build and evaluate RAG pipelines for fare rules and hotel policies; prompt engineering, retrieval quality evaluation, and hallucination monitoring

Voice AI: contribute to inference pipeline for spoken intent understanding; benchmark ASR quality on Indian language test sets

Coupon/Promo ML: build personalised offer targeting models; measure redemption lift vs. discount cost with causal evaluation

Sentiment systems: deploy multi-source review summarisation models; build complaint intent classifiers for customer service routing

Experimentation & Measurement:
Design A/B experiments: pre-registration, power analysis, SRM checks, guardrail metric monitoring, and business interpretation

Apply causal inference methods: CUPED variance reduction, quasi-experimental designs, uplift modelling for targeted interventions

Build experiment playbooks adopted across the AI team your rigor sets the standard

The AI-First Mandate:

- AI is not an enhancement.

- It is the product architecture.

- Every surface, every API, every decision point is either ML-powered today or on the roadmap to be.

- Search & Ranking Learning-to-Rank across flights, hotels, bus routes, train coaches; real-time re-ranking on user signals

- Voice AI Hindi/Hinglish voice booking, intent resolution, spoken fare comparisons, accessibility-first conversational UX

- RAG Systems Fare rule retrieval, hotel cancellation policy Q&A, airline contract intelligence, real-time regulatory updates

- Agentic AI Autonomous booking resolution, exception handling, refund orchestration, supplier communication bots

- MCP Orchestration Model Context Protocol tool chains across GDS APIs, payment gateways, and supplier integrations

- Recommendation Engine Cross-vertical next-best-action, collaborative filtering, session-based deep learning

- Price Intelligence Competitive fare mapping, lower-price guarantee engine, demand elasticity, yield optimisation

- Coupon & Promo ML Personalised offer targeting, redemption probability scoring, margin-aware discount optimisation

- Sentiment & Review AI Review summarisation, NPS prediction, complaint triage, trust signal extraction

- Fraud & Risk ML Anomaly detection, account takeover signals, payment fraud scoring, fake review classification

- Deep System Mapping Route intelligence, geo-semantic matching, multi-modal journey planning

- Predictive Systems Cancellation risk, no-show prediction, seat upgrade probability, waitlist conversion

Who You Are :

- 710 years in Data Science/ML; shipped at least 3 production models with measurable, sustained business impact

- Strong ML fundamentals: supervised/unsupervised learning, gradient boosting, neural networks, NLP, ranking systems

- Production engineering comfort: your pipelines have unit tests, monitoring, and runbooks not just notebooks

- LLM literacy: you have built or evaluated at least one RAG system; you understand chunking, retrieval quality, and output evaluation

- Statistical rigour: you pre-register experiments, check SRM, apply CUPED, and do not celebrate

- Tier-I institute preferred (IIT / IIIT / NIT / IISC / BITS CSE / AI / Statistics)

Technology Stack :

- ML: PyTorch XGBoost LightGBM Scikit-learn Optuna SHAP

- NLP/LLM: HuggingFace LangChain Sentence Transformers spaCy FastEmbed

- Feature Eng: Feast Flink Kafka Streams Spark dbt

- Serving: TorchServe Triton BentoML FastAPI

- Experiment: MLflow Weights & Biases Custom A/B infra