Job Description
ABOUT THE ROLE
Axtria is seeking a technically deep AI Engineer to design and deliver production-grade GenAI and Agentic AI solutions across its pharma and life sciences client portfolio. You will work across the full data and AI stack — medallion architecture, structured and unstructured data pipelines, Informatica IDMC-based integration, and intelligent agent systems — deploying across client environments that span Microsoft, OpenAI, and AWS ecosystems. Platform expertise is client-dependent; candidates are expected to be proficient across all three.
CLIENT-DEPENDENT AI PLATFORMS
Engagements vary by client infrastructure. Candidates must be capable across all three platforms.
Location: Tokyo - Hybrid (3 days onsite)
Microsoft Copilot Studio Copilot build & publish · Custom connectors · Power Platform · Teams deployment · Enterprise governance | ChatGPT Enterprise Enterprise rollout · OpenAI API · GPT-4o · Fine-tuning · Prompt governance · Data privacy controls | AWS Bedrock Foundation model APIs · Agents for Bedrock · Knowledge bases · RAG on S3 · Lambda · Guardrails |
CORE TECHNICAL PILLARS
GenAI LLM integration · RAG · Fine-tuning · Prompt engineering · Evaluation · Vector DBs | Agentic AI Multi-agent systems · LangGraph · AutoGen · Tool use · Memory · Planning loops | Data Engineering Medallion (Bronze/Silver/Gold) · Databricks · Delta Lake · Structured & unstructured data | Data Integration Informatica IDMC · MDM · Data quality · Lineage · Cloud connectors |
KEY RESPONSIBILITIES
- Design and deploy GenAI solutions — RAG pipelines, LLM-powered workflows, and enterprise AI assistants — across Copilot Studio, ChatGPT Enterprise, and AWS Bedrock depending on client infrastructure.
- Build Agentic AI systems using multi-agent frameworks (LangGraph, AutoGen, Agents for Bedrock) enabling autonomous reasoning, tool use, memory, and multi-step planning.
- Develop and maintain medallion architecture (Bronze → Silver → Gold) on Databricks / Delta Lake, handling structured data (CRM, sales, clinical) and unstructured data (documents, PDFs, clinical notes, emails).
- Implement Informatica IDMC pipelines for data ingestion, data quality, MDM, and lineage — integrating cloud and on-prem pharma data sources into unified data layers.
- Build Microsoft Copilot Studio copilots with custom connectors, Power Platform workflows, and Teams-based enterprise deployment for Microsoft-stack clients.
- Deploy ChatGPT Enterprise at scale — including prompt governance frameworks, usage policies, API integrations, and data privacy controls.
- Leverage AWS Bedrock for foundation model APIs, knowledge base-backed RAG, Agents for Bedrock orchestration, and Guardrails for responsible AI.
- Build semantic retrieval and vector search layers (Pinecone, OpenSearch, Azure AI Search, Bedrock Knowledge Bases) to power intelligent agents.
- Define and enforce data governance, cataloguing, and lineage standards through IDMC and Databricks Unity Catalog.
- Engage pharma clients in Japan — conducting requirements workshops, solution demos, and delivery reviews in Japanese.
- Champion responsible AI practices: agent safety, guardrails, model observability, and prompt governance across all platform deployments.
REQUIRED QUALIFICATIONS
- 4–8 years in AI/ML engineering, data engineering, or applied AI — with at least 2 years on GenAI or LLM-based production systems.
- Hands-on experience with at least two of: Microsoft Copilot Studio, ChatGPT Enterprise / OpenAI API, AWS Bedrock — with working familiarity across all three.
- Proven experience designing Agentic AI systems: multi-agent orchestration, tool-use patterns, memory management, and planning with LangGraph, AutoGen, or equivalent.
- Deep proficiency with Databricks (Delta Lake, MLflow, Unity Catalog) and medallion architecture for structured and unstructured data.
- Hands-on Informatica IDMC experience — CDI, CAI, data quality rules, MDM, or data lineage workflows.
- Strong Python skills; proficiency with LangChain, LangGraph, LlamaIndex, or similar AI orchestration frameworks.
- Experience processing unstructured data: document parsing, NLP pipelines, OCR, or embedding generation.
- Business-level Japanese proficiency (JLPT N2 or above) and professional English proficiency.
PREFERRED QUALIFICATIONS
- Working knowledge of pharma functions — commercial operations, R&D, sales force analytics, or medical affairs.
- Experience with pharma data platforms: Veeva CRM, IQVIA, Symphony Health, or equivalent.
- Cloud certifications: AWS ML Specialty / Solutions Architect, Azure AI Engineer, or Databricks certifications.
- Familiarity with Copilot Studio governance, Power Automate flows, and Microsoft 365 Copilot extensibility.
- Experience with AWS SageMaker, Azure OpenAI Service, or Azure Data Factory in production environments.
GENAI & AGENTIC AI SKILLS
ChatGPT Enterprise | OpenAI API / GPT-4o | AWS Bedrock | Copilot Studio |
RAG pipelines | LangGraph | AutoGen | Agents for Bedrock |
Prompt engineering | Vector databases | Multi-agent orchestration | Fine-tuning |
DATA PLATFORM SKILLS
Databricks | Delta Lake | Medallion architecture | Informatica IDMC |
Structured data | Unstructured data | MLflow | MDM / Data lineage |
CLOUD & INTEGRATION
AWS SageMaker | AWS Lambda | Azure OpenAI | Power Platform |
Azure Data Factory | Python | LangChain | Japanese (N2+) |
Axtria is an equal opportunity employer and is committed to creating an inclusive environment for all employees.