Senior Data Architect - AI-Powered Data Platforms

without disrupting current operations.
As a GE Vernova accelerator, GE Vernova Advanced Research is driving strategy and leading research & development efforts to execute on the business's mission to help power the energy transition. We forge the collaborations and help invent the technologies required to electrify and decarbonize for a zero-carbon future.

Representing virtually every major scientific and engineering discipline, our researchers are collaborating with GE Vernova's businesses, the U.S. government, and more than 420 entities at the forefront of technology to execute on 150+ energy-focused projects. Collectively, these research programs and initiatives aim to solve near term technical challenges, deliver next generation product advances, and drive long term breakthrough innovation to enable more affordable, reliable, sustainable, and secure energy.

Job Description

Key job responsibilities

Unstructured Data & AI Enablement

• Design scalable architectures for processing and indexing unstructured data (PDFs, documents, emails, logs, images) for AI consumption

• Architect document processing pipelines that leverage multi-modal LLMs (GPT-4V, Claude, Gemini) for direct document understanding without traditional OCR preprocessing

• Implement intelligent document extraction using LLMs' native vision and context capabilities to handle complex layouts, tables, and mixed media

• Design metadata extraction and enrichment pipelines that enhance discoverability of unstructured assets

• Build architectures for multi-modal AI applications that combine structured and unstructured data sources

RAG & Knowledge Platform Architecture

• Design end-to-end RAG architectures that leverage existing data lakes and enterprise knowledge bases

• Architect hybrid search systems combining traditional keyword search with semantic/vector search capabilities

• Implement chunking strategies and embedding pipelines for diverse document types and data sources

• Build architectures for continuous learning where RAG systems are updated with new data in near real-time

• Design security and access control models that work across legacy systems and modern AI platforms

• Create data governance frameworks that ensure compliance while enabling AI innovation

Platform Optimization & Scale:

• Optimize storage strategies for cost-effective management of structured and unstructured data

• Design tiered storage architectures that balance performance needs with storage costs

• Implement caching layers for frequently accessed embeddings and AI model inputs

QUALIFICATIONS

• Bachelor's degree in Computer Science, Information Systems, or related field

• 10+ years of experience as a Data Architect, Data Platform Engineer, or similar role with enterprise data systems

• 5+ years of experience working with both structured (SQL databases, data warehouses) and unstructured data (documents, logs, multimedia)

• Understanding of modern document processing using multi-modal LLMs and traditional extraction methods

• Proficiency in Python and SQL, with experience in data processing libraries

• Must be willing to work out of an office located in Bangalore JFWTC Campus

• You must submit your application for employment on the careers page at www.careers.gevernova.com to be considered.

PREFERRED QUALIFICATIONS

• 12+ years of experience modernizing legacy data architectures for cloud and AI workloads

• Deep expertise in unstructured data processing using both multi-modal LLMs and traditional methods

• Experience with multi-modal LLMs for document understanding and their cost/performance trade-offs

• Background in information retrieval, search engineering, or content management systems

• Experience with multi-modal AI architectures combining text, image, and structured data

• Master's degree in Computer Science, Information Systems, or related field

Technical Stack

Document Processing: Multi-modal LLMs (GPT-4V, Claude Vision, Gemini), LlamaParse, Unstructured.io, Azure Document Intelligence, AWS Textract (for legacy/high-volume), direct PDF-to-context pipelines

Vector/Search: Pinecone, Weaviate, pgvector

Lake Technologies: AWS S3, Azure ADLS

Languages: Python, SQL, Scala, Java

APIs: OpenAI, Anthropic, Google Vertex AI, AWS Bedrock, Azure OpenAI

Additional Information

Relocation Assistance Provided: Yes