Forward Deployed Engineer, Enterprise - Milan

Murphy AI; Lombardia; On-site; EUR 90,000 - 140,000; Full time

Job summary

Murphy AI in Milan seeks an experienced Forward Deployed Engineer to deliver enterprise AI agent deployments for banks. You will scope with customer tech teams, design effective agent interactions, and launch the solution on client systems across channels.

You’ll own production performance, analyze data, and drive improvements in agent behavior and integration with core banking and data pipelines.

Qualifications

• Experience delivering enterprise AI deployments for financial services.
• Ability to scope, design and launch agent-based solutions in production.
• Strong SQL and data interpretation skills.

Responsibilities

• Own end-to-end enterprise agent builds from scoping to production launch.
• Define agent behavior, guardrails and escalations with customers.
• Design prompts and logic for multi-language, multi-channel conversations.

Skills

End-to-end delivery

POC scoping

Agent design

SQL

Tools

SQL

APIs

Data pipelines

Telephony

Job description

Murphy deploys AI agents for banks starting with the most straightforward business case: voice AI for debt collection. We’ve built a fully-fledged platform designed to optimize recovery rates while maintaining respectful and personalized communication.

Our AI-powered agents adapt instantly, engaging with debtors across channels like voice, messengers, email and SMS to maximize results while preserving trust.

We’re live with Santander, Revolut, BBVA and dozens others. We’ve proven the product in large-scale pilots with top-tier banks. And now we need to move fast to secure leadership in this $300Bn+ industry.

Collections agents only work in production when someone owns the build in detail. Each one handles regulated conversations, sensitive customer situations and payment outcomes, and has to behave correctly every time.

Forward Deployed Engineers own that build. They turn a customer's collections strategy, policies and systems into a working agent, prove it in a POC, and take it to full production. They work with Account Executives during the sale, with Deployment Strategists after go-live, and with Product and Engineering to make the platform better with every deployment.

As a Forward Deployed Engineer, you own the technical execution of AI agent builds for our enterprise customers, mainly banks and large lenders. You scope with the customer's technical and operations leads, design the agent, build and test it, and launch it on their systems. After launch, you own the agent's performance: you analyze production data, explain results to the customer, and decide what to change in the agent and the contact strategy. Each deployment should leave behind patterns, tools and tests that make the next one faster.

You’ll be based in Milan and lead enterprise deployments for banks and large lenders in Italy.

In this role, you will

• Own end-to-end delivery of enterprise agent builds, from POC scoping through implementation, evaluation and production launch.
• Define the agent operating procedures (actions, guardrails, escalations) with the customer, and turn them into agent behavior.
• Design and engineer the prompt and agent logic that drives identification, negotiation and payment conversations, across languages and channels.
• Configure Murphy to fit each client’s needs on how and when to reach debtors: call windows, retry and frequency rules within each country's regulations, channel mix (voice, SMS, RCS, WhatsApp), payment links and hand-offs to human agents.
• Build evaluation and simulation harnesses, plus regression tests against real call outcomes, to catch errors and drift before customers do.
• Build, validate and maintain integrations with customer systems: core banking, collections platforms, data pipelines, telephony.
• Diagnose failures from transcripts, traces and logs, find the root cause, and fix it in the agent or flag it to Engineering.
• Analyze production performance (contact rates, promises, payments, funnel) with SQL and dashboards, explain changes