Data Scientist

Servify; Mumbai; On-site; INR 1,200,000 - 1,800,000; Full time

Job summary

Servify, a global product lifecycle platform, is hiring a Data Scientist to own predictive analytics within our data ecosystem. With 2–4 years in data science and Databricks production experience, you will maintain ML pipelines, ensure data quality, and drive platform transformation.

You'll work with PySpark, Python, SQL, and feature engineering across PostgreSQL and MongoDB, integrating outputs into Databricks SQL Analytics and visualization platforms to support strategic decisions.

Qualifications

• 2-4 years of building, evaluating, and managing robust ML models and predictive infra.
• Databricks: production experience in managing analytics jobs, compute clusters, and data feature assets.
• Proficiency in Python (PySpark/Pandas/Scikit-Learn) and advanced SQL; comfortable with Databricks notebooks and MLflow.
• Database proficiency to query large datasets from PostgreSQL and MongoDB for training data.
• Experience with Medallion Architecture data engineering for feature extraction.
• BI tool exposure (Tableau or Databricks SQL Analytics/Warehouse) to map metrics.

Responsibilities

• Model stabilization, monitoring, and reliability of production Databricks models and pipelines.
• Maintain and iterate feature extraction processes across Bronze to Gold layers.
• Lead end-to-end integration of ML outputs onto Databricks visualization platforms and warehouses.
• Design training data ingestion for diverse sources (PostgreSQL, MongoDB).
• Share and secure ML outputs and visualizations with internal and external stakeholders.
• Tune cloud compute costs and model scoring latency by optimizing clusters and Delta structures.

Skills

Python (PySpark/Pandas/Scikit-Learn)

SQL

Databricks notebooks

MLflow

Feature Engineering

Tableau

Tools

Databricks

PostgreSQL

MongoDB

Tableau

Job description

Servify is a global product lifecycle management platform transforming the after-sales service experience. Founded in 2015 and headquartered in India, Servify connects leading OEM brands with their sales and service ecosystems to deliver seamless customer support and service operations. Servify partners with leading brands, retailers, distributors, insurers, service providers, and carriers across Asia, North America, Europe, and the MENA region.

Servify embraces and values diversity and is committed to be an equal opportunity employer. We welcome applications from all members and will give equal consideration to all irrespective of age, disability, gender, sexual orientation, ethnicity, nationality, religion or belief.

Why Join Servify?

At Servify, you'll be part of a fast-growing global technology company that encourages innovation, ownership, and continuous learning. We offer competitive compensation, ESOPs, employee-friendly policies, modern collaboration tools, professional development opportunities, and benefits that help you do your best work.

Position Summary: We are seeking an experienced Data Scientist to take operational ownership of predictive analytics and algorithmic models within our established data ecosystem. If you have 2-4 years of data science experience and Databricks in a production environment, this is your chance to drive a critical platform transformation. This role provides significant autonomy. You will be responsible for the continuous health of our production machine learning pipelines, ensuring data quality, statistical model fidelity, and leading the charge to leverage advanced

analytical assets directly inside Databricks SQL Analytics and Warehouse to support intelligent business decisions.

Key Responsibilities:

• Model Stabilization & Management: Manage, monitor, and ensure the 24/7 operational reliability and accuracy of our current suite of production Databricks (PySpark/Delta Lake) predictive models and analytical pipelines. You are directly responsible for model operational health.
• Data & Feature Fidelity: Maintain and iterate on complex feature extraction processes structured around the Medallion Architecture. Guarantee the statistical integrity and performance of analytical data features moving from Bronze to Gold layers
• Predictive Insights Lead: Execute the end-to-end integration of machine learning outputs and predictive logic onto the native Databricks visualization platform and SQL Analytics warehouses. This is a core focus area.
• Source System Integration for Analytics: Design and optimize training data ingestion logic for diverse data sources, with specific responsibility for extracting, transforming, and loading analytical training sets efficiently from PostgreSQL and MongoDB.
• Partner Insights Sharing & Security: Establish and govern secure, reliable mechanisms for sharing finalized Databricks machine learning outputs, data visualizations, and predictive metrics with both internal stakeholders and external partners.
• Cost & Model Performance Optimization: Actively tune distributed machine learning tasks, cluster configurations, and Delta table structures to drive down cloud computing costs and minimize model scoring latency.

Requirements:

• 2-4 years' experience in building, evaluating, and managing robust machine learning models and predictive infrastructure.
• Databricks: Recent hands-on production experience in managing analytics jobs, compute clusters, and data feature assets within the Databricks environment.
• Expert proficiency in Python (PySpark/Pandas/Scikit-Learn) and advanced SQL. You should be comfortable working extensively with Databricks notebooks and ML flow terminals.
• Database Expertise: Proven ability to connect to, query, and efficiently extract large datasets from PostgreSQL (Postgres DB) & MongoDB to generate clean structures for analytical model training (Understanding of NoSQL schema and extraction methods is key).
• Architecture & Methodology: Practical experience working with data engineered under the Medallion Data Architecture to pull features cleanly for predictive scoring.
• BI Tool Knowledge: Prior exposure to dashboards (such as Tableau or Databricks SQL Analytics/Warehouse) is essential for mapping historical metrics to forward-looking predictive metrics.