Machine Learning Engineer

TE-4 Years and above

Location- Bangalore/Chennai/Hyderabad

NP- 15-30 Days max

Join our fast‑growing team to build a unified platform for data analytics, machine learning, and generative AI. You’ll integrate the AI/ML toolkit, real‑time streaming into a backed feature store, and dashboards—turning raw events into reliable features, insights, and user‑facing analytics at scale.

What you’ll do

• Design and build streaming data pipelines (exactly‑once or effectively‑once) from event sources into low‑latency feature serving and NRT and OLAP queries.
• Develop an AI/ML toolkit: reusable libraries, SDKs, and CLIs for data ingestion, feature engineering, model training, evaluation, and deployment.
• Stand up and optimize a production feature store (schemas, SCD handling, point‑in‑time correctness, TTL/compaction, backfills).
• Expose features and analytics via well‑designed APIs/Services; integrate with model serving and retrieval for ML/GenAI use cases.
• Build and operationalize Superset dashboards for monitoring data quality, pipeline health, feature drift, model performance, and business KPIs.
• Implement governance and reliability: data contracts, schema evolution, lineage, observability, alerting, and cost controls.
• Partner with UI/UX, data science, and backend teams to ship end‑to‑end workflows from data capture to real‑time inference and decisioning.
• Drive performance: benchmark and tune distributed DB (partitions, indexes, compression, merge settings), streaming frameworks, and query patterns.
• Automate with CI/CD, infrastructure‑as‑code, and reproducible environments for quick, safe releases.

Tech you may use

Languages: Python, Java/Scala, SQL

Streaming/Compute: Kafka (or Pulsar), Spark, Flink, Beam

Storage/OLAP: ClickHouse (primary), object storage (S3/GCS), Parquet/Iceberg/Delta

Orchestration/Workflow: Airflow, dbt (for transformations), Makefiles/Poetry/pipenv

ML/MLOps: MLflow/Weights & Biases, KServe/Seldon, Feast/custom feature store patterns, vector stores (optional)

Dashboards/BI: Superset (plugins, theming), Grafana for ops

Platform: Kubernetes, Docker, Terraform, GitHub Actions/GitLab CI, Prometheus/OpenTelemetry

Cloud: AWS/GCP/Azure

What we’re looking for

1. 4+ years building production data/ML or streaming systems with high TPS and large data volumes.
2. Strong coding skills in Python and one of Java/Scala; solid SQL and data modeling.
3. Hands‑on experience with Kafka (or similar), Spark/Flink, and OLAP stores—ideally ClickHouse.
4. GenAI pipelines: retrieval‑augmented generation (RAG), embeddings, prompt/tooling workflows, model evaluation at scale.
5. Proven experience designing feature pipelines with point‑in‑time correctness and backfills; understanding of online/offline consistency.
6. Experience instrumenting Superset dashboards tied to ClickHouse for operational and product analytics.
7. Fluency with CI/CD, containerization, Kubernetes, and infrastructure‑as‑code.
8. Solid grasp of distributed systems and architecture fundamentals: partitioning, consistency, idempotency, retries, batching vs. streaming, and cost/perf trade‑offs.
9. Excellent collaboration skills; ability to work cross‑functionally with DS/ML, product, and UI/UX.
10. Ability to pass a CodeSignal prescreen coding test.

Grid Dynamics (Nasdaq:GDYN) is a digital-native technology services provider that accelerates growth and bolsters competitive advantage for Fortune 1000 companies. Grid Dynamics provides digital transformation consulting and implementation services in omnichannel customer experience, big data analytics, search, artificial intelligence, cloud migration, and application modernization. Grid Dynamics achieves high speed-to-market, quality, and efficiency by using technology accelerators, an agile delivery culture, and its pool of global engineering talent. Founded in 2006, Grid Dynamics is headquartered in Silicon Valley with offices across the US, UK, Netherlands, Mexico, India, Central and Eastern Europe.

www.griddynamics.com

--