AI Solution Engineer

EPAM Systems; United States; Remote; USD 180,000 - 260,000; Full time

Job summary

EPAM Systems Inc. is seeking a Lead to architect and deliver AI-native software teams. You will own the delivery process, guide problem framing, and collaborate directly with client stakeholders to ensure high-quality results in agentic development projects.

You will set up agent repos, design SDLC stages, and establish guardrails for safe AI usage while mentoring engineers and contributing to PoCs and presales. Extensive Java/.NET/Python experience is required.

Qualifications

• 7+ years in software engineering with hands-on Java, .NET or Python
• 6+ months of daily, hands-on use of AI coding agents on real projects
• Strong design and debugging skills and experience with agentic development

Responsibilities

• Lead technical delivery of a small AI-native team from framing to production
• Define the team's agentic SDLC and set up repos for agents
• Define how AI-generated code is checked with TDD, test automation, and CI/CD quality gates
• Set safe-use rules for agents including permissions, secrets, data handling
• Measure and optimize AI cost and delivery impact (tokens, model choice, cycle time)
• Own the solution architecture and mentor engineers; contribute to PoCs and presales

Skills

Java

.NET

Python

Code debugging

AI coding agents

Tools

GitHub Copilot agent mode

CLI

Claude Code

Codex

Job description

We're forming small AI-native teams where AI coding agents are the default way to build software, whether that's web apps, APIs, data pipelines, modernization or agentic systems for clients. This role is about engineering with agents, not only building them. As Lead, you define how the team works with agents, own delivery quality, and work directly with client stakeholders.

Responsibilities

Lead technical delivery of a small AI-native team, from problem framing and specs to production Design and evolve the team's agentic SDLC: spec-driven workflow, agent configuration, parallel and background agent work, review and quality gates
Set up repos for agents: instructions, prompt files, skills, custom agents, MCP servers, plus the architecture, docs and tests that let agents work reliably
Define how AI-generated code is checked: TDD, test automation, code review, static and security analysis in CI/CD
Set safe-use rules for agents: permissions, secrets, client data, approved tools
Measure and optimize AI cost and delivery impact (tokens and premium requests, model choice, cycle time, rework)
Own the solution architecture; mentor engineers in agentic practices; contribute to PoCs, estimates and presales

Requirements

7+ years in software engineering, with hands-on Java, .NET or Python, and strong design and debugging skills 6+ months of daily, hands-on use of AI coding agents (GitHub Copilot agent mode / CLI / coding agent, Claude Code or Codex) on real projects Deep context engineering for coding agents: repo instructions, prompt files, skills, custom agents and subagents, MCP Spec-driven development in practice, including breaking work into tasks an agent can do well Proven practices for keeping AI-generated code at production quality: TDD, test automation, reviewing large AI diffs, CI/CD quality gates Understanding of agent security risks: permissions, secrets, data leakage, prompt injection through repos and tools Practical AI cost awareness: model selection per task, token and premium-request economics, measuring productivity impact Solution architecture skills; experience leading and mentoring engineers and working with clients English level B2+ Nice to have Experience building LLM or agentic applications (agent frameworks, RAG, tool calling, evals, guardrails, LLM observability) Showcase of rolling out AI-assisted development to other teams (enablement, playbooks, metrics) Cloud and AI certifications (Azure, AWS, GCP, GitHub Copilot, Anthropic) Business analysis mindset with consulting or presales experience Familiarity with UX/UI and the financial services domain