Senior AI Software Engineer

DNVChennai, Tamil NaduOn-siteFull-timeSenior, 5–8 yearsListed 18 minutes ago

Apply now

About this role

The Senior AI Software Engineer designs, develops, and maintains production-grade AI engines, APIs, services, and reusable platform enhancements across Digital & Transformation. Sitting within the Data & AI Engineering section, the role builds the AI capabilities that power the broader platform, ensuring they are scalable, reusable, secure, and able to support multiple customer implementations and business workflows. The work helps advance how DNV performs Due Diligence, Verification & Assurance, and Renewables Certification across Energy Systems.

An AI engine here is a reusable, API-exposed service that packages LLM orchestration, prompts, retrieval, and extraction logic so it can serve multiple products and customer implementations.

You will work in close partnership with Product leadership, Solution Engineering, Application Engineering, Platform Reliability, and Solution Architecture, translating product needs, implementation feedback, and complex workflow requirements into reliable, maintainable, production-ready solutions. The focus is on intelligent document processing and extraction, LLM-powered workflows, Retrieval-Augmented Generation (RAG), and agentic AI that can understand, process, and generate structured information from complex documents.

Python and hands-on AI engine development are the primary technical requirements. Experience with LangChain, LangGraph, and other AI orchestration frameworks is highly relevant. JavaScript and Node.js experience is a plus.

This role is based at our DNV office in Chennai, India. Further details regarding role-specific requirements will be shared during the interview process.

Key Responsibilities

AI Engines, LLM Orchestration, RAG & Agentic AI

- Design, develop, test, and maintain scalable AI engines that provide reusable intelligence and automation across multiple platform implementations.
- Build modular, LLM-powered architectures using LangChain, LangGraph, or comparable orchestration frameworks, supporting multiple models, prompts, agents, tools, and workflows.
- Integrate and manage LLM providers such as OpenAI, Azure OpenAI, and other relevant model providers, with model selection, fallback, token management, context handling, and cost optimization.
- Design and implement agentic workflows involving multi-step reasoning, tool calling, function calling, and task orchestration.
- Design and implement RAG architectures using vector databases and embedding models for semantic search, contextual retrieval, and knowledge-based applications.
- Build AI-powered search, chatbot, document query, and intelligent reporting capabilities, with conversation memory and context management.

Intelligent Document Processing & Extraction

- Build end-to-end document intelligence pipelines covering ingestion, preprocessing, classification, extraction, validation, and structured output generation.
- Develop advanced, document-aware extraction workflows using LLMs and prompt engineering for complex, unstructured documents in formats such as PDF, DOCX, and spreadsheets (scanned documents are a plus).
- Implement context-aware extraction across pages, sections, tables, and related documents, supporting single-document, multi-document, and cross-document scenarios.
- Handle complex structures, missing information, ambiguous content, and varying layouts, with confidence scoring, validation rules, and exception handling.
- Develop reusable extraction templates, prompt configurations, and document-specific processing strategies.
- Optimize extraction accuracy, processing time, token consumption, and overall AI execution cost.

Backend Engineering, APIs & Platform Integration

- Develop scalable backend services, APIs, and microservices that expose AI engine capabilities to enterprise applications, primarily using Python, FastAPI, Pydantic.
- Design asynchronous processing, event-driven workflows, streaming responses, and background processing for long-running AI operations.
- Establish reusable API contracts, shared services, and integration patterns that ensure interoperability across platform components.
- Integrate AI engines with document management systems, databases, cloud storage, and enterprise applications.
- Implement secure document access, tenant isolation, authorization, and data-handling mechanisms across AI workflows.
- Work with relational and NoSQL databases, object storage, and vector databases, keeping services modular, scalable, and independently deployable.

AI Quality, Evaluation & Optimization

- Develop evaluation strategies and testing frameworks to measure AI output quality, extraction accuracy, and model performance.
- Validate structured AI outputs against predefined schemas, business rules, and expected results.
- Implement prompt evaluation, model comparison, regression testing, and continuous improvement workflows.
- Analyze failed extractions, retrieval inaccuracies, hallucinations, and inconsistent responses, and optimize prompts, retrieval strategies, model configurations, and orchestration accordingly.
- Build reusable evaluation datasets, test scenarios, and quality benchmarks.
- Ensure AI solutions are explainable, traceable, measurable, and suitable for production.

Engineering Practices, Reliability & Security

- Use AI-assisted development practices to accelerate coding, testing, documentation, refactoring, and analysis, while maintaining strong code quality, security, and review standards.
- Contribute to CI/CD pipelines, automated testing, deployment automation, and production monitoring.
- Build reliable AI services with appropriate logging, observability, error handling, retry and timeout mechanisms, rate limiting, and fault tolerance.
- Follow secure development practices aligned with internal security standards, SOC 2 Type II, ISO 27001, and enterprise expectations, with attention to data handling, access control, auditability, and responsible AI.
- Take ownership of AI services from architecture and development through deployment and production support.

Collaboration, Technical Ownership & Leadership

- Work effectively with Product, Solution Engineering, Application Engineering, Platform Reliability, Solution Architecture, and other stakeholders to translate complex business requirements into practical AI architectures and implementation approaches.
- Review feature requests from Solution Engineering and help turn requirements into reusable product or platform enhancements.
- Collaborate with Solution Engineering on complex document types, failed extractions, prompt optimization, and reusable implementation patterns.
- Own the technical decisions for the components you build, and drive complex challenges independently from design through production support.
- Lead design and code reviews to ensure consistency, maintainability, and engineering quality.
- Mentor and guide other engineers on AI engine architecture, prompt engineering, document extraction, and integration patterns.

Support knowledge transfer to Solution Engineering and Application Engineering, and contribute to technical standards, reusable frameworks, architecture documentation, and engineering best practices.

- Bachelor's or Master's degree in Computer Science, Engineering, Data Science, Information Systems, or a related field, or equivalent practical experience.
- 5+ years of professional software engineering experience, with substantial hands-on experience developing AI-enabled applications, backend services, AI engines, or intelligent platforms.
- Strong Python programming skills, with hands-on experience designing and developing production-grade AI engines and LLM-powered applications.
- Hands-on experience building AI engines with LangChain, LangGraph, or comparable AI orchestration frameworks.
- Strong understanding of LLMs, prompt engineering, model integration, embeddings, tokenization, context windows, and structured outputs.
- Proven experience developing intelligent document processing and extraction solutions using LLMs, including ingestion, preprocessing, classification, extraction, validation, and structured data generation.
- Practical experience with RAG architectures, vector search, and semantic retrieval, and with building AI agents, multi-step workflows, and tool or function calling.
- Experience building backend services, REST APIs, and asynchronous processing using Python and FastAPI or equivalent technologies, with relational or NoSQL databases and cloud storage.
- Understanding of AI evaluation methodologies, extraction accuracy measurement, and prompt optimization.
- Experience with automated testing, CI/CD, source control, and code review, and a strong grasp of software architecture, security, scalability, and maintainability.
- Ability to independently troubleshoot complex AI engineering challenges and deliver production-ready solutions.

What Is Preferred

- Experience building AI engines or reusable AI platforms from the ground up.
- Experience with document intelligence platforms, intelligent document processing (IDP), OCR, document parsing tools, and complex data extraction.
- Familiarity with multimodal AI, vision-language models, and processing scanned or image-based documents.
- Experience with Azure OpenAI, OpenAI, Anthropic, or other LLM providers.
- Experience with vector databases such as Pinecone, MongoDB, or similar technologies.
- Experience with cloud-native AI deployments on Azure, and with Docker, Kubernetes, serverless architectures, and distributed processing.
- Experience with modern UI frameworks such as React or comparable frontend frameworks.
- Familiarity with JavaScript and Node.js for backend development, API integrations, and platform interoperability.
- Background in energy, infrastructure, renewables, due diligence, verification, assurance, certification, or other technically complex industries.
- Experience developing reusable prompt libraries, AI evaluation frameworks, and enterprise AI engineering standards.
Security and compliance with statutory requirements in the countries in which we operate is essential for DNV. Background checks will be conducted on all final candidates as part of the offer process, in accordance with applicable country-specific laws and practices.