Senior AI-Native Solutions Architect

EPAM Systems; United States; Remote; USD 160,000 - 230,000; Full time

Benefits offered by this job

Medical, Dental and Vision Insurance (

401(k) Plan with Matching

Paid Time Off

Job summary

EPAM Systems, Inc. seeks an AI Architect to design AI-native applications and enterprise AI solutions. You will define orchestrations, retrieval, integration, and guardrails while shaping business-aligned decisions and architecture documents.

You will prototype to provide evidence and coordinate with delivery teams across client engagements. You will ensure clear accountability, cost models, and observability, guiding both engineers and coding agents to implement safe, effective AI-native

Qualifications

• Experience architecting AI-native systems in production (RAG, assistants, agents, agentic workflows).
• Ability to surface trade-offs with business terms and owners.
• Proven delivery accountability for complex enterprise architectures.

Responsibilities

• Design end-to-end AI-native architectures including orchestration, retrieval, and tool integration.
• Define evaluations, guardrails, and constraints for agents and autonomous components.
• Document architecture decisions and trade-offs for future teams to reuse.
• Collaborate with delivery teams and presales to shape proposals and client options.
• Prototype where evidence is needed to guide decisions.

Skills

AI-native architectures

Distributed systems

Stakeholder communication

Architecture decision records

Prototyping

Tools

OAuth 2.1

SPIFFE

AuthZEN

DPoP

Job description

EPAM Systems, Inc. seeks an AI Architect to design AI-native applications and enterprise AI solutions. You will define orchestrations, retrieval, integration, and guardrails while shaping business-aligned decisions and architecture documents.

You will prototype to provide evidence and coordinate with delivery teams across client engagements. You will ensure clear accountability, cost models, and observability, guiding both engineers and coding agents to implement safe, effective AI-native