AI Engineer
Industry: Large pharmaceutical industry
Location: Doha, Qatar
Employment Type: Full-time, on-site in office
Company: Intelligence Experts, Doha, Qatar
POSITION OVERVIEW
We are seeking an exceptional AI Engineer to design, build, and evolve sophisticated multi-agent AI systems for a large pharmaceutical industry environment. This role focuses on production-grade agentic AI, manufacturing intelligence, quality control, and operational analytics across complex pharmaceutical operations.
The AI Engineer will work hands-on with modern agentic frameworks including LangGraph, LangChain, and LangFuse, while integrating enterprise data infrastructure such as Azure/AWS, Snowflake, Neo4j, vector databases, and full-stack Python and React applications.
Impact: Direct influence on systems that optimize batch processing, enable real-time anomaly detection, and synthesize insights from billions of data points across pharmaceutical manufacturing operations.
KEY RESPONSIBILITIES
1. Agentic Architecture and Design
- Design and evolve the core agentic architecture supporting multi-agent workflows, including planning, data fetching, synthesis, analysis, and reporting.
- Define state management patterns, checkpoint strategies, and memory systems for long-running agent conversations.
- Architect Human-in-the-Loop integration patterns for quality assurance and risk mitigation.
- Establish best practices for agent composition, tool design, and inter-agent communication.
- Create technical roadmaps that balance innovation with production stability.
2. Hands-On Development
- Write production-quality Python code for critical agent components.
- Build LangGraph state management, checkpoint services, and session handling.
- Develop data agents for SQL query generation, semantic validation, document parsing, embedding, and knowledge graph traversal.
- Develop orchestration agents for task planning, dependency management, and workflow coordination.
- Build analysis agents for visualization generation, anomaly detection, and ML-driven insights.
- Implement sophisticated prompt engineering for SQL generation, synthesis, and reasoning tasks.
- Build robust validation pipelines, including SQL injection prevention, schema validation, and result sanity checks.
- Develop real-time monitoring and observability instrumentation using LangFuse.
- Build and maintain full-stack features using Python backend services and React frontend interfaces.
3. Framework and Stack Expertise
- Support adoption and optimization of LangGraph, LangChain, LangFuse, and Deep Agents.
- Work with LangGraph for multi-agent state machines, graph-based workflows, and parallel execution patterns.
- Work with LangChain for tool definitions, chains, retrieval-augmented generation, and agent workflows.
- Work with LangFuse for agent tracing, observability, and performance analytics.
- Apply advanced agentic patterns including reflection, planning, and tool-use optimization.
- Maintain deep knowledge of emerging agentic frameworks and contribute to technology evaluation.
- Guide technology choices, including when to use LLMs vs. SLMs, caching strategies, and cost optimization.
4. Cloud and Data Infrastructure
- Design and implement integrations with Snowflake, Neo4j, ChromaDB, Azure AI services, and AWS AI stack.
- Work with Snowflake for query optimization, cost control, and schema design.
- Work with Neo4j for semantic search, relationship modeling, and document discovery.
- Work with vector stores such as ChromaDB, Pinecone, or Weaviate for embedding management, semantic indexing, and RAG optimization.
- Architect file system abstraction layers for Azure Blob Storage, S3, and local storage.
- Design and optimize database schemas for checkpoint persistence and result tracking.
- Implement connection pooling, caching strategies, and performance optimization.
5. Collaboration and Knowledge Sharing
- Collaborate with AI engineers, data engineers, ML researchers, manufacturing teams, and business stakeholders.
- Conduct code reviews with attention to architectural consistency and quality.
- Pair program on complex implementations, including prompt engineering, agent coordination, and validation.
- Share knowledge through documentation, architecture decision records, and technical discussions.
- Contribute to engineering practices, including testing strategies, deployment procedures, and incident response.
6. Quality, Testing, and Reliability
- Design comprehensive validation frameworks.
- Build unit tests for agent components with mocked LLM responses.
- Build integration tests for multi-agent workflows.
- Build end-to-end tests simulating real manufacturing queries.
- Implement safety guardrails including SQL injection prevention, query cost estimation, and anomaly detection.
- Establish error handling and graceful degradation patterns.
- Drive observability through structured logging, distributed tracing, and performance dashboards.
7. Production Operations and Optimization
- Manage and improve deployment pipelines using Docker, Kubernetes, and CI/CD automation.
- Monitor system health, latency, and cost metrics post-launch.
- Implement observability dashboards using LangFuse, Prometheus, and custom analytics.
- Optimize performance through LLM caching, query batching, and result streaming.
- Support incident response for agent failures, data quality issues, and system degradation.
PREFERRED QUALIFICATIONS
Research and Innovation
- Familiarity with academic agentic AI research papers.
- Contributions to AI/ML open-source projects.
- Participation in AI communities, research communities, or technical forums.
Domain Experience
- Manufacturing, supply chain, pharmaceutical, or quality control domain knowledge.
- Experience with anomaly detection or time-series analysis.
- Knowledge of data quality frameworks and validation patterns.
Advanced Skills
- Machine learning model development and evaluation.
- NLP and embeddings experience, including Hugging Face and sentence-transformers.
- Real-time data processing with Kafka, Flink, or similar technologies.
- Database performance tuning and query optimization.
WHAT WE ARE BUILDING
The AI Engineer will help build a sophisticated multi-agent agentic system for a large pharmaceutical industry environment.
System Architecture
The system includes 10 major components across 12 execution phases:
- Input Processing: Session management and request validation.
- Memory and State: LangGraph-based state store with checkpoint persistence.
- Planning: Task decomposition and dependency management.
- Data Agents: SQL query generation, document retrieval, and knowledge graph traversal.
- Human-in-the-Loop: Data quality validation before synthesis.
- Synthesis: Multi-source data consolidation and context building.
- Analysis Planning: Task breakdown for visualization and ML analysis.
- Visualization Agent: Chart and dashboard generation.
- ML Analysis Agent: Anomaly detection and predictive insights.
- Report Assembly: Final insight synthesis and formatting.
- Checkpoint Persistence: Session state archival.
- Suggested Questions: Dynamic follow-up generation.
Technical Stack
- Core: Python 3.10+, FastAPI, Pydantic, React.
- Frontend: React, JavaScript or TypeScript, data dashboards, API-driven user interfaces.
- Agentic Frameworks: LangGraph, LangChain, LangFuse, Deep Agents.
- Data: Snowflake, Neo4j, ChromaDB, PostgreSQL for checkpoints, Redis for caching.
- Cloud: Azure AI or AWS Bedrock, Blob Storage, and compute services.
- DevOps: Docker, Kubernetes, GitHub Actions or GitLab CI.
- Monitoring: LangFuse, Prometheus, and structured logging.
Business Context
- Operations in a large pharmaceutical industry environment, including batch processing, quality control, and supply chain visibility.
- Structured and unstructured data analysis, combining SQL queries with document insights.
- Real-time anomaly detection and proactive alerting.
- User-facing intelligent assistant capabilities for manufacturing teams asking complex questions and receiving synthesized answers.
WHY THIS ROLE IS UNIQUE
- Technical breadth and depth: Work across systems thinking, hands-on coding, Python, React, full-stack implementation, and agentic AI.
- Emerging technology: Work with cutting-edge agentic frameworks before they become mainstream.
- Real-world impact: Build systems that support manufacturing decisions in a large pharmaceutical industry environment.
- Collaborative engineering: Work closely with AI engineers, data teams, and domain specialists on production-grade systems.
- Innovation culture: Help establish practical best practices in agentic AI systems for pharmaceutical operations.
COMPETENCIES AND MINDSET
Technical Mindset
- Systems thinker: Understand how components interact and anticipate failure modes.
- Pragmatist: Choose appropriate trade-offs between simplicity, performance, and extensibility.
- Continuous learner: Comfortable keeping pace with the rapidly evolving agentic AI landscape.
- Quality-first: Prioritize reliability and robustness for production systems.
Interpersonal Skills
- Communicator: Explain complex agentic architectures to non-technical stakeholders.
- Collaborator: Work cross-functionally with data engineers, ML researchers, manufacturing teams, and business stakeholders.
- Ownership: Take accountability for implementation quality and system outcomes.
- Knowledge sharer: Contribute to documentation, reviews, and engineering standards.
