About Us
Foundever™ is a global leader in the customer experience (CX) industry. With 150,000 associates across the globe, we are the team behind the best experiences for more than 750 of the world's leading and digital-first brands. Our innovative CX solutions, technology, and expertise are designed to support operational needs for our clients and deliver seamless, AI-driven experiences in the moments that matter. As AI becomes increasingly agentic and autonomous, we are building a platform to orchestrate intelligent workflows that automate complex business processes at scale.
Job Summary
We are looking for an Agentic Technical Lead to own the agentic systems framework — the platform and tooling that enables technical teams across the organization to build, configure, evaluate, and deploy agentic systems for their own use cases.You will architect and ship the infrastructure that makes agentic development self-service and production-safe: configuration interfaces where developers define system prompts and models for each agent component, end-to-end evaluation pipelines with predefined metrics, dataset creation and experiment management via LangFuse, and iterative workflows that take teams from prototype to production. You will define how MCP Servers and Clients are templated via Backstage, ensuring governance and visibility over all deployed agentic systems. You will continuously expand the framework's capabilities so teams can build increasingly sophisticated agents without reinventing infrastructure.The stack includes LangGraph for orchestration, LangFuse for observability and evaluation, and AWS infrastructure for scale. This role requires deep expertise in LLMs and agentic systems, strong architectural thinking, and the ability to lead end-to-end in a fast-moving AI environment.
Primary Job Responsibilities
Platform & Framework
Architect the agentic systems framework that other teams use to build, configure, and deploy agents
Build configuration interfaces (e.g., chat-based UIs) for defining system prompts, selecting models, and composing agent topologies
Implement MCP Server and Client templates via Backstage for standardized scaffolding and catalog visibility
Continuously improve the framework — new agent patterns, tool integrations, orchestration abstractions, and reusable components
Evaluation & Testing Infrastructure
Build end-to-end evaluation pipelines with predefined quantitative metrics (accuracy, latency, cost, task completion) and qualitative assessments (coherence, safety, user satisfaction)
Enable dataset creation in LangFuse, LLM-as-judge pipelines, and experiment management workflows so teams can configure, evaluate, iterate, and ship with confidence
Define reference evaluation standards that teams can use out of the box and extend for their use cases
Observability & Production
Implement observability via LangFuse — execution tracing, cost tracking, quality drift monitoring across all deployed agents
Design dashboards and alerting for performance, anomalies, and drift detection
Own production reliability of the framework and its core components
Technical Leadership
Own the technical vision and roadmap for the platform
Lead design decisions from prototype through production, including release cycles and rollback strategies
Mentor engineers on agentic patterns, evaluation practices, and framework usage
Build reference agentic systems that serve as templates for adopting teams
Collaboration
Work with adopting teams to onboard them, understand their needs, and feed requirements into the platform roadmap
Collaborate with ML, data, product, and DevOps teams to ensure the framework meets real needs at scale
Stay current with advances in agentic AI, evaluation methodology, and developer tooling
Skills / Abilities / Knowledge
Experience
8+ years in machine learning engineering or applied ML
4+ years hands-on with LLMs (fine-tuning, prompt engineering, integration, deployment)
2+ years building and shipping agentic systems in production, end-to-end
Proven experience building platforms or tooling used by other engineering teams
Deep experience with evaluation frameworks — dataset creation, metric definition, LLM-as-judge implementations
Required
Strong proficiency in Python
Hands-on experience with LangGraph, LangFuse, and MCP Servers/Clients
Deep understanding of LLM integration patterns: tool calling, structured outputs, prompt chaining, RAG
Strong evaluation methodology knowledge for generative AI and agentic systems
Experience with relational databases (PostgreSQL or similar)
Excellent debugging and system-level thinking across multi-step agent executions
Comfortable in Agile, fast-paced environments with evolving requirements
Nice to have
AWS (compute, storage, networking, IAM)
Vector databases (Pinecone, Weaviate, Qdrant, pgvector)
Graph databases (Neo4j or similar)
Product-facing system experience — shipping features, measuring impact
NLP background — particularly speech-to-text, transcription, or diarization
Backstage or similar internal developer portals
Kubernetes and CI/CD pipelines
Skills
Education
Languages
Excellent written and conversational English (C1 minimum)
French and Spanish are a plus
Education
Tools & Technologies
Our offer
Impactful work: Lead the platform enabling agentic AI adoption across a global organization
Professional growth: Work at the frontier of agentic systems with continuous learning opportunities
Competitive compensation: Attractive salary and benefits package
Collaborative environment: Remote-friendly team with opportunities for travel, training, and industry events
Foundever is a global leader in the customer experience (CX) industry with 170,000 associates across the globe serving more than 750 of the world's leading and digital-first brands. The company provides innovative CX solutions, technology, and expertise designed to support operational needs for clients and deliver seamless customer experiences. Foundever operates significant office locations including a Sofia center employing over 800 people and supporting 32 languages worldwide. The company specializes in contact center operations, customer service delivery, and virtual care services including healthcare advisory and telehealth support.
