Data Scientist

Who we are:

devx Labs is an AI-native consulting company that fundamentally reimagines business operations around AI capabilities. We don’t just implement AI tools - we transform how enterprises operate through first-principles thinking and outcome-driven execution.

About the role:

• You own the modelling layer of client engagements: turning customer, transaction and operational data into propensity, value, demand and pricing models that drive decisions across Customer Interactions and AI-Led Business Operations.
• Clients span retail, D2C, telecom, financial services and manufacturing. The problem changes every few months; the discipline doesn't: frame the question, prove the signal exists, ship a model that holds in production, show the business the number moved.
• Not a notebook-only role. You are accountable for the model from EDA through deployment, monitoring and the A/B read that proves it worked.

Problems you'll work on

• Customer Lifetime Value and customer health scoring
• Churn prediction, retention and win-back propensity
• Purchase, upsell and cross-sell propensity; Next Best Action
• Customer segmentation, behavioural clustering and persona generation
• Demand intelligence: SKU, store and channel forecasting, promo and seasonality effects
• Pricing and promotion optimisation: price elasticity, markdown and discount depth
• Uplift modelling and campaign targeting
• Anomaly detection and fraud or risk scoring
• Marketing effectiveness and attribution
• Recommendation and personalisation for commerce and CX

Build and Ship Models - 60%

• Frame the problem:translate a business question into a modelling target, label definition, unit of analysis and success metric before writing code.
• Own the data work:EDA, data quality assessment, feature engineering and leakage checks on large transactional and behavioural datasets in SQL and Python.
• Build the model: select and train the right approach across classical ML (GBMs, GLMs, survival models, clustering), time-series forecasting and deep learning where it earns its place (sequence models, embeddings).
• Evaluate honestly: out-of-time validation, calibration, lift and gain curves, business-metric backtests, error analysis by segment. A good AUC on a leaky split is a failure.
• Deploy and operate: package models for batch scoring or API serving on Vertex AI or SageMaker, track experiments (MLflow or equivalent), monitor drift, data quality and performance decay in production.
• Prove impact: design and read A/B tests, uplift tests and holdouts to measure incremental business value, not just model accuracy.
• Explain the model: SHAP, partial dependence and segment-level diagnostics that make outputs defensible to business and risk stakeholders.

Drive Client Success - 30%

• Shape the approach: participate in discovery workshops, assess data readiness and define the analytics approach with Practice Leads and Solutions Consultants.
• Present with clarity: explain model results to client business and technology teams: what it does, where it is uncertain, how to act on it.
• Integrate: work embedded with client data and engineering teams to land models in their CRM, marketing automation, planning or pricing workflows.
• Iterate: track adoption and business outcomes post-launch; recalibrate as the business changes.

Multiply the Practice - 10%

• Build accelerators: feature libraries, evaluation templates, model cards and playbooks for repeat problems such as churn, CLV and demand.
• Mentor: guide DS 1 team members and interns on modelling rigour, code quality and client communication (DS 2 expectation).
• Champion AI-assisted development: use Claude Code, Cursor or similar to accelerate EDA, pipeline and test writing while owning correctness.

The Scientist Mindset

• Rigour first: you distrust a result until you have checked the split, the leakage and the baseline.
• Outcome obsessed: success is retained customers, forecast error reduced and margin gained, not model count.
• Pragmatic: you know when a well-tuned GBM beats a transformer, and when it doesn't.

Technical Depth

• Statistics and ML fundamentals: regression, classification, clustering, time-series, hypothesis testing, causal inference basics (uplift, DiD, propensity matching). You can explain why a technique works, not just how to call it.
• Production modelling experience: hands-on with at least two of: CLV, churn, propensity, segmentation, demand forecasting, pricing or uplift. Kaggle-only or coursework-only experience doesn't clear the bar.
• Python and SQL: pandas, scikit-learn, XGBoost / LightGBM / CatBoost, statsmodels; strong SQL against cloud warehouses (BigQuery, Redshift, Snowflake or Databricks).
• Deep learning: working knowledge of PyTorch or TensorFlow for tabular, sequence or embedding use cases; judgement on when DL is warranted.
• Evaluation and testing: cross-validation strategy design, calibration, backtesting, unit tests for feature pipelines, validation of model outputs before release.
• Deployment and MLOps: has taken a model into production (batch or real-time) with experiment tracking, versioning and monitoring. Exposure to Vertex AI, SageMaker, MLflow, Airflow or equivalent.
• Data at scale: PySpark or Dask for large datasets; comfort with messy, multi-source enterprise data.
• Explainability and responsible AI: SHAP / LIME, fairness checks and clear model documentation.

Good to have

• LLM APIs or embeddings for feature enrichment, text analytics or agentic workflows.
• Optimisation and prescriptive methods: linear / integer programming, multi-armed bandits, reinforcement learning basics for pricing or allocation.
• Marketing mix modelling or Bayesian methods.
• Domain exposure to retail, D2C, telecom, BFSI or manufacturing.
• GCP Professional ML Engineer or AWS ML Specialty certification.

Client-Facing Excellence

• Data storytelling: turn a confusion matrix into a business decision a CMO can act on.
• Workshop contributor: comfortable in discovery sessions, data assessments and solution reviews with client teams.
• Collaborative: works effectively with consultants, data engineers, AI engineers and client analysts.

Experience and Background

• DS 1: 2 to 3 years of hands-on data science experience with at least one model shipped to production.
• DS 2: 3 to 5 years of hands-on data science experience with multiple models shipped and owned end to end, including post-launch monitoring and iteration, plus some mentoring or project ownership.
• Education: Bachelor's or Master's in Statistics, Mathematics, Computer Science, Economics, Engineering or a related quantitative field.
• Stakeholder experience: consulting, product companies or startups where you worked directly with business stakeholders.
• Portfolio: GitHub repos, case studies or write-ups you can walk us through.

• AI tool credits and freedom to experiment with modern approaches.
• Performance-based appraisal at 6 months + annual appraisal cycle.
• Competitive pay with a structured CTC breakup.
• Work with clients across industries and domains.
• Hybrid model with work-from-home support and flexible working hours.
• No-cap leave policy — take time off responsibly, no fixed limits.
• Health insurance with spouse and kids coverage included.
• Yearly outings, and an open environment for innovation.