Meghana Reddy Ganugapenta
I am a senior AI/ML Data Engineer who designs and delivers scalable data and ML platforms across healthcare, finance, and retail.
About
I am a senior AI/ML Data Engineer who designs and delivers scalable data and ML platforms across healthcare, finance, and retail. I specialize in end-to-end LLM-based NLP and RAG systems, using models like BERT, BioBERT, Flan-T5, GPT-4, and LLaMA-2, together with LangChain, Pinecone, and FAISS to enable high-precision semantic search, summarization, and context-aware retrieval. My work spans data engineering, MLOps, and governance to produce reliable, production-grade AI solutions that inform strategic decisions.
I enjoy architecting lakehouses and modern data ecosystems across AWS, Azure, and GCP, ensuring reproducible training, real-time inference, and continuous monitoring of data quality and embedding performance. I thrive collaborating with cross-functional teams to operationalize AI responsibly, maintain compliance, and deliver actionable insights through dashboards and ML-driven metrics.
- 10 years' experience
- Joined January 2026
Industry experience
- Healthcare
- Financial Services
- Retail
- Government
- Education
More experts in Los Angeles, United States
Work experience· 5 roles
Designed and built scalable ETL/ELT pipelines using Azure Data Factory, Databricks, and PySpark to ingest health care data from FHIR/HL7 APIs, lab systems, databases, and IoT sources. Automated training-data pipelines for LLM and RAG workflows, including chunking, embedding generation, metadata enrichment, and dataset versioning. Implemented semantic search and entity-linking platforms with Pinecone, Azure Cognitive Search, and FAISS for cross-dataset medical entity retrieval. Applied NLP and transformer models (BERT, BioBERT, ClinicalBERT, DeBERTa-v3) to extract clinical entities and relationships from unstructured text. Developed generative AI solutions with GPT-4 and AWS Bedrock (LLaMA-2, Claude) for summarization and clinical reporting. Standardized RAG orchestration and observability with LangGraph, LangSmith, and MCP, improving prompt tuning and explainability. Implemented LLM safety/governance controls including structured prompting, validation tests, and embedding-based consist
Designed and operated cloud-native data pipelines on GCP using BigQuery, Dataflow, and Cloud Composer to consolidate housing, health, and community data. Implemented event-driven ingestion with Pub/Sub and Cloud Functions for near real-time synchronization across agency platforms. Standardized and transformed complex datasets using MongoDB aggregations and PostgreSQL, flattening nested JSON. Established secure, schema-governed workflows with RBAC, ensuring HIPAA compliance. Enabled end-to-end observability and lineage via Cloud Monitoring, Data Catalog, and MLflow; tracked pipeline latency, schema changes, and health metrics. Orchestrated ETL and ML workloads with Airflow/Cloud Composer; automated dependency management, refresh cycles, and retraining triggers. Built curated analytical layers in BigQuery with partitioning and clustering for cost-effective queries. Enforced data quality validation with Great Expectations and Python checks. Developed predictive modeling pipelines with Ten
Modernized legacy on-prem and Oracle data platforms by migrating to Azure Data Lake Gen2 and Azure Synapse, enabling elastic scalability and governance. Built high-performance lakehouse pipelines with Delta Lake, Databricks, and Azure Data Factory, leveraging Photon acceleration and Unity Catalog for optimized processing and access control. Implemented event-driven data workflows with Azure Functions and Event Grid; automated downstream refreshes and alerts. Established batch and streaming ingestion using Apache NiFi, Kafka, Azure Event Hubs, and Databricks Structured Streaming. Organized Bronze/Silver/Gold lakehouse layers with Delta Lake, applying Z-ordering, partitioning, and time travel for analytics and auditability. Tuned PySpark and SQL workloads to reduce compute costs. Developed feature-engineering pipelines for forecasting, demand planning, and churn modeling. Applied time-series modeling in Python and PySpark to improve forecast accuracy. Managed ML experiments with MLflow;
Education
Hire Meghana Reddy Ganugapenta today
Post a job and invite Meghana, or contact them directly. No hidden commission.






