Job description:
Experienced AWS Agentic AI Engineer to design, build, and deploy autonomous and semi‑autonomous AI agents on AWS. This role focuses on developing goal‑driven, tool‑using, multi‑step AI systems capable of reasoning, planning, executing actions, and collaborating with humans and enterprise systems to solve real‑world business problems.
The position sits at the intersection of Large Language Models (LLMs), cloud‑native architecture, APIs, and automation, delivering production‑grade agentic AI solutions using AWS AI/ML and cloud services.
Experience & Qualifications
- Bachelor’s degree in computer science, Engineering, Technology, or a related field, with 8–12 years of overall IT experience and at least 3+ years of hands‑on experience building agentic AI solutions using AWS cloud‑native services.
- Strong expertise in core AWS services including AWS Lambda, Step Functions, EventBridge, S3, DynamoDB or Aurora, OpenSearch, Amazon Bedrock, and Amazon SageMaker.
- Advanced proficiency in Python for building scalable AI systems and automation workflows.
- Proven experience leading the technical design and implementation of complex AI agent architectures and components.
- Hands‑on experience designing Retrieval Augmented Generation (RAG) solutions and applying model fine‑tuning techniques.
- Demonstrated ability to manage performance, scalability, reliability, and cost optimization of AI workloads in production environments.
- Proven experience conducting code reviews, leading knowledge‑sharing sessions, and making critical decisions on technologies, architectures, and frameworks.
- Healthcare domain experience is desirable, along with knowledge of AI safety, governance, compliance, and responsible AI practices.
- AWS certifications such as Solutions Architect or Machine Learning Specialty are preferred, and experience with LangChain and LangGraph is a plus.
- Excellent communication skills with a strong ability to collaborate effectively with both technical and non‑technical stakeholders.
Roles & Responsibilities
- Design and develop agentic AI systems using Amazon Bedrock and other foundation models, including Claude and Titan.
- Build autonomous and semi‑autonomous AI agents capable of multi‑step reasoning, planning, tool usage, action execution, and human‑in‑the‑loop collaboration.
- Integrate with Amazon SageMaker for custom model training, fine‑tuning, evaluation, and experimentation.
- Implement Retrieval Augmented Generation (RAG) solutions using Amazon OpenSearch, Amazon Aurora, DynamoDB, and vector databases such as FAISS or Pinecone.
- Optimize inference cost, latency, scalability, and reliability across AI workloads.
- Implement guardrails, validation layers, and human‑in‑the‑loop controls to ensure safe, reliable, and predictable AI behavior.
- Address hallucination mitigation, prompt injection risks, bias concerns, and model misuse scenarios.
- Ensure compliance with enterprise AI governance, security, and regulatory standards.
- Implement comprehensive logging, monitoring, observability, and audit trails for AI systems.
- Support production deployments, incident resolution, and root cause analysis for AI services.
- Continuously improve agent performance, reliability, robustness, and usability through iteration, monitoring, and feedback.
- Collaborate closely with architecture, DevOps, data engineering, security, and business teams to deliver end‑to‑end AI solutions.
Over Cognizant
Cognizant (NASDAQ: CTSH) is een bouwer van AI-oplossingen en een leverancier van technologiediensten. Wij slaan de brug tussen AI-investeringen en ondernemingswaarde door het bouwen van full-stack AI-oplossingen voor onze klanten. Onze diepgaande kennis van sectoren, processen en engineering stelt ons in staat om de unieke context van een organisatie te verankeren in technologische systemen. Deze systemen versterken het menselijk potentieel, realiseren tastbare resultaten en geven wereldwijde ondernemingen een voorsprong in een snel veranderende wereld. Ontdek hoe op cognizant.ai of @cognizant.
Aanvullende arbeidsinformatie
De informatie over de beloning is correct op de datum van deze vacature. Cognizant behoudt zich het recht voor om deze informatie op elk moment te wijzigen, met inachtneming van de toepasselijke wetgeving.
Van sollicitanten kan worden verwacht dat zij gesprekken bijwonen, persoonlijk of via een videogesprek. Daarnaast kan van kandidaten worden gevraagd om tijdens elk gesprek een geldig, door de overheid uitgegeven identiteitsbewijs (zoals een identiteitskaart of paspoort) te tonen.
Cognizant is een werkgever die gelijke kansen biedt. Je sollicitatie en kandidatuur worden niet beoordeeld op basis van ras, huidskleur, geslacht, religie, levensovertuiging, seksuele geaardheid, genderidentiteit, nationale afkomst, handicap, genetische informatie, zwangerschap, veteranenstatus of enige andere eigenschap die wordt beschermd door federale, regionale of lokale wetgeving.
