Data Engineer (Python/AI) - Assistant Vice President
Citi · Pune · 8+ yrs experience · Posted 2026-07-18
Tech stack: AWS, Docker, FastAPI, Flask, Kubernetes, Linux, Python, Redis
Apply on the company site · Get a referral for this role
Citi salary & ratings · Citi interview process · More live openings
About the role
We are looking for a mid-level
Python Developer
with combined experience in
Data Engineering
and
AI/NLP engineering
. The candidate will build NLP pipelines using libraries such as
Flair
BERT
, and LLM frameworks, and will also work on large-scale data processing using
PySpark
Pandas
, and related data tools. The role includes developing APIs, integrating with platform services, and supporting CI/CD deployments using GitHub and LightSpeed Enterprise.
Responsibilities: - Develop and optimize ETL/data processing jobs using PySpark Pandas PyArrow and related libraries.
- Work with Parquet files using FastParquet pyarrow.parquet for efficient data processing.
- Implement data parsing and serialization using json ujson or orjson for high-performance JSON handling.
- Build and maintain NLP pipelines using Flair, BERT, and LLM-based models
- Develop scalable ingestion and data transformation pipelines for AI and analytics use cases.
- Build and maintain Flask-based APIs for model inference and service integrations.
- Use regular expressions for text cleaning, parsing, and NLP preprocessing.
- Integrate caching and fast lookups using Redis
- Manage and deploy ML models using MLflow for tracking and versioning.
- Support CI/CD workflows using GitHub LightSpeed Enterprise and deployment pipelines.
- Create and maintain Autosys JILs for job scheduling and automation.
- Use basic Linux commands for troubleshooting, operations, and deployment tasks.
- Monitor application and system health using ITRS Geneos
- Write unit tests and improve automation test coverage ( PyTest /unittest).
- Work with REST APIs, microservices, and basic shell scripting.
- Work with cloud services (ECS), including boto3 Required Skills
- 8+ years of hands-on Python programming experience.
- Strong fundamentals in Python, OOP, and design patterns.
Qualifications: - with NLP libraries such as Flair, BERT, HuggingFace Transformers, or similar.
- Solid experience with PySpark Pandas PyArrow and distributed data pipelines.
- Proficient in working with Parquet using FastParquet pyarrow.parquet
- Familiarity with fast JSON parsing libraries (json, ujson, orjson).
- building APIs using Flask (FastAPI is a plus).
- with MLflow for model tracking and deployment.
- Good understanding of CI/CD practices and Git workflows.
- working with Redis or similar in-memory stores.
- with Autosys JILs for job scheduling.
- Comfortable with Linux command line and shell scripting.
- Strong debugging, problem-solving, and teamwork skills.
- Exposure to cloud services; AWS boto3 is an asset.
- Nice-to-Have with Polars or Dask for high-performance data processing.
- with PyTorch or TensorFlow for model training.
- with Docker, Kubernetes, or containerized deployments.
- with monitoring tools such as ITRS Geneos.
- with FastAPI, Airflow, or Prefect.
Qualifications
- with NLP libraries such as Flair, BERT, HuggingFace Transformers, or similar.
- Solid experience with PySpark Pandas PyArrow and distributed data pipelines.
- Proficient in working with Parquet using FastParquet pyarrow.parquet
- Familiarity with fast JSON parsing libraries (json, ujson, orjson).
- building APIs using Flask (FastAPI is a plus).
- with MLflow for model tracking and deployment.
- Good understanding of CI/CD practices and Git workflows.
- working with Redis or similar in-memory stores.
- with Autosys JILs for job scheduling.
- Comfortable with Linux command line and shell scripting.
- Strong debugging, problem-solving, and teamwork skills.
- Exposure to cloud services; AWS boto3 is an asset.
- Nice-to-Have with Polars or Dask for high-performance data processing.
- with PyTorch or TensorFlow for model training.
- with Docker, Kubernetes, or containerized deployments.
- with monitoring tools such as ITRS Geneos.
- with FastAPI, Airflow, or Prefect.
Responsibilities
- Develop and optimize ETL/data processing jobs using PySpark Pandas PyArrow and related libraries.
- Work with Parquet files using FastParquet pyarrow.parquet for efficient data processing.
- Implement data parsing and serialization using json ujson or orjson for high-performance JSON handling.
- Build and maintain NLP pipelines using Flair, BERT, and LLM-based models
- Develop scalable ingestion and data transformation pipelines for AI and analytics use cases.
- Build and maintain Flask-based APIs for model inference and service integrations.
- Use regular expressions for text cleaning, parsing, and NLP preprocessing.
- Integrate caching and fast lookups using Redis
- Manage and deploy ML models using MLflow for tracking and versioning.
- Support CI/CD workflows using GitHub LightSpeed Enterprise and deployment pipelines.
- Create and maintain Autosys JILs for job scheduling and automation.
- Use basic Linux commands for troubleshooting, operations, and deployment tasks.
- Monitor application and system health using ITRS Geneos
- Write unit tests and improve automation test coverage ( PyTest /unittest).
- Work with REST APIs, microservices, and basic shell scripting.
- Work with cloud services (ECS), including boto3 Required Skills
- 8+ years of hands-on Python programming experience.
- Strong fundamentals in Python, OOP, and design patterns.