EPAM Systems
1 day ago
Lead AI Engineer — TCO Agent Platform
Sign up to save this job, get alerts, and apply with an optimized CV.
Company information
- Company
- EPAM Systems
- Location
- USA United States
- Posted
- 1 day ago
Job description
As the Lead AI Engineer for our next-generation TCO Agent Platform, you will serve as the technical lead and multi-agent system architect for a greenfield, proactive FinOps AI platform. You will oversee the development, orchestration, and governance of an ecosystem comprising 8 specialized AI agents (e.g., Commitment Optimization, Financial Operations, Resource Optimization, Anomaly Detection) running within a multi-cloud GCP/AWS environment.
You will drive core architectural execution, ensure strict adherence to SOX-adjacent financial controls, enforce Zero Trust security models via Model Armor and LiteLLM, and guide the engineering team through building high-scale, autonomous enterprise AI services.
Reporting directly to the Technical Product Manager, you will collaborate closely with Solution, Platform, and Enterprise Architects. You will also have a Senior Software Engineer under your direct subordination to collaborate with on solution implementation.
Req.#1087619193
Responsibilities
- Multi-Agent Architecture & Orchestration: Lead the design and implementation of 8 specialized agents using Python, FastMCP, and GCP Workload Identity. Oversee inter-agent dependencies, prompt engineering lifecycle, tool definitions, and agent-to-service communication
- LLM Governance & Tokenomics: Enforce centralized LLM routing via LiteLLM Gateway and Vertex AI Model Garden (Claude, Gemini). Implement agent self-governance tracking systems (Tokenomics) to monitor and cap LLM operating costs within strict platform operational limits
- Financial & Compliance Guardrails: Architect execution boundaries and strict segregation-of-duties workflows for SOX-adjacent processes (e.g., Journal Entry generation vs. human approval)
- System Integration & Action Routing: Oversee the architecture of the platform’s Action Gateway—handling direct API invocations, event-driven workflows, and fallback ticketing (Jira, Slack, Teams)
- Technical Leadership & Standards: Set coding, testing, and formatting standards across application repositories. Mentor Senior and Mid-level AI engineers and drive code reviews enforcing RFC standard error formats, API contracts, and schema compliance
Requirements
- Experience: 8+ years of software engineering experience with 3+ years in a technical leadership capacity building multi-agent AI systems, FinOps tools, or LLM-powered platforms
- Frameworks & Languages: Advanced proficiency in Python 3.11+, FastMCP, FastAPI, and Pydantic. Prior experience with object-oriented enterprise languages (e.g., Java) for seamless integration with core platform services and backend APIs
- AI/LLM Architecture: Hands-on experience with Google ADK, Vertex AI, LiteLLM Gateway, Model Armor guardrails, prompt engineering, structured tool output parsing, and agent execution boundaries
- Data & Cloud Platforms: Deep familiarity with the GCP ecosystem (BigQuery, GKE, Workload Identity), SQL schema design (FOCUS standard preferred), and partitioned/clustered OLAP architectures
- Security & Governance: Experience implementing Zero Trust authentication (OAuth/KSA-to-GSA), RBAC, immutable audit logging, and API/MCP error specifications (RFC 7807/9457)
- DevOps & Infrastructure: Proficiency with Docker builds, GKE deployment patterns, OpenTofu/Terraform, and CI/CD pipelines
- Leadership Skills: Proven ability to coach, mentor, and influence teams beyond just writing and implementing solutions
Nice to have
- Experience with LangGraph / LangChain
- Hands-on experience with Vertex AI
- Working knowledge of modern DevOps and CI/CD practices
- Familiarity with GCP infrastructure resources and their constraints, with the ability to identify optimal resources for designed AI agentic solutions
Required skills
- development
- java
- security
- governance
- jira
- python
- aws
- prompt engineering
- software engineering
- standards
- leadership skills
- gcp
- claude
- devops
- terraform
- code reviews
- infrastructure
- technical leadership
- architect
- cloud platforms
- proactive
- ci/cd pipelines
- ai agents
- gemini
- slack
- fastapi
- teams
- vertex ai
- frameworks
- devops practices
- financial operations
- oauth
- system integration
- langgraph
- bigquery
- autonomous
- resource optimization
- senior software engineer
- ci/cd practices
- multi-cloud
- finops
- languages
- technical product manager
- orchestration
- langchain
- anomaly detection
- ai engineer
- coach
- financial controls
- mentor
- ecosystem
- gke
- google adk
- rbac
- coding standards
- ai platform
- solution architects
- senior ai engineers
- greenfield
- engineering team
- backend apis
- opentofu
- enterprise architects
- data platforms
- pydantic
- api contracts
- testing standards
- multi-agent architecture
- litellm
- tokenomics
- event-driven workflows
- model armor
- multi-agent system
- fastmcp
- high-scale
- workload identity
- multi-agent ai systems
- llm routing
- formatting standards
- finops tools
- platform architects
- tco agent platform
- commitment optimization
- architectural execution
- sox-adjacent
- zero trust security models
- enterprise ai services
- gcp workload identity
- inter-agent dependencies
- tool definitions
- agent-to-service communication
- llm governance
- litellm gateway
- vertex ai model garden
- agent self-governance tracking
- llm operating costs
- platform operational limits
- financial guardrails
- compliance guardrails
- execution boundaries
- segregation-of-duties
- sox-adjacent processes
- journal entry generation
- human approval
- action routing
- action gateway
- api invocations
- fallback ticketing
- application repositories
- mid-level ai engineers
- rfc standard error formats
- schema compliance
- llm-powered platforms
- python 3.11+
- object-oriented enterprise languages
- core platform services
- ai/llm architecture
- model armor guardrails
- structured tool output parsing
- agent execution boundaries
- gcp ecosystem
- sql schema design
- focus standard
- partitioned olap
- clustered olap
- zero trust authentication
- ksa-to-gsa
- immutable audit logging
- api error specifications
- mcp error specifications
- rfc 7807
- rfc 9457
- docker builds
- gke deployment patterns
- influence teams
- gcp infrastructure resources
- ai agentic solutions
Interested in this position?
Create your free account and tailor your CV to match this job.