Do you have a passion for working with Australia-based Deloitte Tax clients and helping shape the next generation of AI-enabled tax operations? Are you ready to apply your expertise in Python, LLM applications, RAG, agent orchestration, and AI evaluation to solve complex Tax challenges? Do you want to work with Tax specialists, engineers, and product teams to transform Tax processes into secure, explainable, reliable, and testable AI solutions?
If the answer is “yes,” join the AU Tax Technology and Transformation group in Deloitte India (Offices of the U.S.), a service line of Deloitte Tax LLP.
At Deloitte, we help clients transform Tax by combining deep Tax expertise with data, automation, and artificial intelligence. Our teams design and operate AI-enabled platforms, agentic workflows, rules engines, and data solutions that improve how Tax work is performed. By bringing together Tax specialists, software engineers, data engineers, and AI technologists, we help clients manage complexity, improve quality, and build scalable and intelligent Tax operations.
At Deloitte, we are leading clients through the tax transformation taking place in the marketplace. We offer a broad range of fully integrated tax services and add greater impact to clients by combining technology and tax technical resources to uncover insights and smarter solutions for navigating an increasingly complex global environment.
Our Vision:
To be the leading global provider of Tax services by delivering exceptional value to our clients and people through trusted relationships, AI-enabled solutions, continuous innovation, and responsible technology adoption.
Role Summary:
We are seeking an Applied AI Engineer to build secure, reliable, and testable AI-assisted and agentic capabilities for next-generation Tax products. The ideal candidate will combine Python, LLM applications, RAG, deterministic rules, structured data, and human review to deliver explainable and production-ready AI solutions.
The role involves close collaboration with Tax specialists, engineers, product teams, and business stakeholders to transform complex Tax processes into scalable AI-enabled features. Experience with AI evaluation, cloud deployment, APIs, data workflows, and responsible AI practices is highly valued.
Key Job Responsibilities:
- Map Tax workflows and determine where to apply AI, deterministic rules, automation, and human review.
- Build and maintain LLM assistants, document-intelligence solutions, and agentic automation capabilities for Tax.
- Develop agent orchestration using LLM APIs, tool calling, structured outputs, state management, and multi-agent design patterns.
- Design and implement RAG pipelines covering document ingestion, chunking, embeddings, vector search, retrieval, grounding, and source management.
- Build AI evaluation frameworks using reference datasets, automated tests, regression testing, quality metrics, monitoring, and feedback loops.
- Monitor AI solutions for quality, latency, cost, reliability, security, and production performance. Implement guardrails, grounding, access controls, traceability, human review, and responsible AI controls to reduce hallucinations and improve trust.
- Combine probabilistic AI with deterministic rules, validation logic, business controls, and auditability requirements. Build, deploy, and support secure AI services across cloud platforms, backend systems, and enterprise environments.
- Translate complex Tax requirements into reliable AI-enabled product features in collaboration with Tax specialists and engineering teams.
- Contribute to reusable AI frameworks, evaluation harnesses, agent tooling, technical documentation, and engineering standards.
- Participate in code reviews, automated testing, release validation, Git, and CI/CD practices to improve solution quality and reliability.
- Diagnose and resolve production issues across AI services, data pipelines, integrations, and Tax product modules.
- Communicate technical decisions, AI outputs, risks, limitations, and solution recommendations to technical and non-technical stakeholders.
- Strong hands-on experience with Python and production software, data, or applied-AI development.
- Practical experience building LLM applications, including prompting, structured outputs, tool calling, model integration, and response validation.
- Experience designing and implementing RAG solutions, including embeddings, vector search, retrieval pipelines, grounding, and source management.
- Experience implementing multi-agent systems using patterns such as ReAct, self-reflection, plan-and-execute, or hierarchical delegation.
- Ability to combine probabilistic AI with deterministic rules, business controls, validation logic, auditability, and human-in-the-loop review.
- Experience evaluating AI systems using reference datasets, automated tests, regression testing, quality metrics, monitoring, and feedback loops.
- Strong understanding of APIs, structured data, backend services, data-processing workflows, and system integration.
- Knowledge of secure software deployment, including cloud hosting, identity and access management, data protection, logging, monitoring, and governance.
- Familiarity with Git, CI/CD, source control, automated testing, code reviews, and modern software delivery practices.
- Ability to translate complex domain requirements into practical product features and technical designs.
- Strong analytical, problem-solving, documentation, communication, and stakeholder-management skills.
- Specific skill set required: Python, LLM applications, RAG, agent orchestration, tool calling, structured outputs, AI evaluation, secure cloud deployment, APIs, data workflows, CI/CD, and automated testing
Work experience: 3 - 6 years
Additional Requirements:
- Experience delivering AI or software products in production environments with requirements for reliability, scalability, security, and operational support.
- Ability to collaborate effectively with Tax specialists, software engineers, data engineers, product managers, and business stakeholders.
- Strong understanding of responsible AI principles, including grounding, guardrails, traceability, access controls, evaluation, and human oversight.
- Exposure to AI research, postgraduate study, technical experimentation, or demonstrated awareness of emerging AI engineering practices.
- Experience with data science and machine-learning tools such as Pandas, scikit-learn, PyTorch, or TensorFlow is advantageous.
- Familiarity with agentic developer tools such as GitHub Copilot and Claude Code is preferred.
- Exposure to Azure OpenAI, Azure AI Foundry, vector databases, Docker, Kubernetes, or MLflow is advantageous.
- Experience with JavaScript, TypeScript, or another relevant programming language is valuable.
- Experience working in Agile delivery environments, including sprint planning, iterative development, testing, reviews, and continuous improvement.
- Strong willingness to learn Tax processes, enterprise architecture, AI technologies, and modern engineering practices.
Work Location - Hyderabad || Shift Timings - 8:00 AM to 5:00 PM IST