About this role
Enterprise customers in Southern Europe struggle to deploy complex AI and global communication solutions on legacy infrastructure. This role exists to solve that challenge by embedding directly with clients in Milan, Italy, and across the region to architect, build, and ship production systems. As a Forward Deployed Engineer on a dedicated Enterprise Sales Pod, the person in this position will handle the entire technical lifecycle from discovery and proofs-of-concept to production go-live.
Working alongside an Enterprise Account Executive, the chosen engineer will own technical execution without relying on a larger local support team. Responsibilities include deploying AI voice assistants, configuring LLM gateways like LiteLLM, managing open-weight models, and setting up robust observability stacks. An ideal candidate combines strong software engineering fundamentals with the communication skills required to lead technical workshops, troubleshoot integration issues, and present directly to enterprise executives in both English and Italian.
What you will do
- Embed directly with enterprise customers to analyze communications workflows, AI use cases, and integration challenges.
- Build and deploy custom implementations utilizing Telnyx APIs, WebRTC, and AI Voice Assistants.
- Run and govern open-weight large language models in customer environments using LiteLLM and various serving engines.
- Wire comprehensive production observability including metrics, logs, traces, dashboards, and alerting systems.
- Design model routing strategies that balance latency, cost, and quality for real-time voice workloads.
- Adapt models to specific regional domains by implementing prompt and RAG pipelines for multilingual use cases and EU regulatory environments.
- Lead proofs-of-concept, pilots, and production launches from initial whiteboard design to stable go-live.
- Create clear technical documentation and runbooks while troubleshooting complex integration issues alongside customer teams.
What you need
- Hold a Computer Science degree or possess equivalent practical experience.
- Bring at least three years of experience building and shipping production software that real users depend on.
- Demonstrate proficiency in multiple programming languages such as Python, Node.js or TypeScript, and Go.
- Show a practical understanding of production AI model failure modes, concurrency, rate limits, and token accounting.
- Feel comfortable deploying containerized services on Kubernetes with proper secrets management and upgrades.
- Possess hands-on experience wiring observability stacks like Prometheus, Grafana, OpenTelemetry, or ELK.
- Provide excellent written and verbal communication skills in both English and Italian for customer-facing technical discussions.
- Be based in Milan, willing to relocate, and legally authorized to work in Italy or eligible for sponsorship.
Nice to have
- Experience with AI voice assistants, STT/TTS, or LLM-based conversational systems
- Building eval sets for a specific domain
- Hands-on production experience with an LLM gateway such as LiteLLM, Portkey, Kong AI Gateway, or an in-house OpenAI-compatible proxy
Skills and tech stack
Questions candidates ask
What is the remote and travel policy for this role?
This is a hybrid role based in Milan, Italy, with travel across Italy, Spain, and Greece. Candidates must be based in Milan or willing to relocate.
Who will I be working with on a day-to-day basis?
You will work side-by-side with a dedicated Enterprise Account Executive as part of a local Enterprise Sales Pod, partnering closely to build the enterprise business in your market.
What kind of technical stack and AI tools will I work with?
You will work with Telnyx APIs, WebRTC, open-weight LLMs like Llama and Mistral, containerized services on Kubernetes, and LLM gateways such as LiteLLM for routing and governance.
What are the core duties regarding customer engagements?
You own the technical side of the engagement, including technical discovery, architecture design, running proofs-of-concept, deploying production workloads, and ensuring long-term solution stability.