Data Engineer
AI Solution Engineer Lead - Milan
Milanfulltimelead
Откликнуться на сайте работодателяОписание вакансии
If you are here, it is because you know that we are looking for an
AI Solutions Engineer Lead
.
Since 2016 we have been using AI to transform the way large companies talk with their customers. The best stories start with great conversations: the relationships that matter are built on the ability to communicate, listen and understand each other. Between companies and people, though, this kind of relationship is still rare, and we work to bring the magic back into interactions, at scale.
Through our AI Agents platform we have automated millions of conversations across voice, chat and messaging, integrating with our clients' CCaaS and enterprise systems and improving their operational and commercial performance. We are a fast-growing European scale-up, backed by Azimut (€15M investment).
What are we looking for?
We are looking for an AI Solutions Engineer Lead: the person who takes AI Agents from design to production with our most complex clients, defines how this craft is done at indigo.ai, and leads the small team doing it (3–4 people today, and growing).
You will start on our most complex enterprise engagement: a multi-country program spanning Europe and Latin America, the proving ground where the method you define gets tested with real users and real traffic, and the natural base from which to grow our wider AI implementation organization.
This is a horizontal, multidisciplinary role sitting between technology and business, not a pure software developer position. You remain the reference architect on the projects where the risk is high and the solution does not exist yet.
The core of the role is functional, architectural and experiential: your value is imagining how an assistant should work (the flows, the integrations, the behavior) and making it actually work the way you imagined it. Your leverage is not typing production code line by line: the deepest technical execution is handled by Product and by the AI tools you use every day. What remains yours, and cannot be delegated, is judgment: how to frame the problem, where it can break, how to make it reliable. And what sets this role apart is that you also need to know how to transfer that judgment to your team.
This is a position with real room to grow: we are scaling our AI implementation organization all over the world, and the scope of this role grows with it: more projects, a larger team, new geographies.
Key Responsibilities:
- Design the agentic architecture. You decide how to frame the problem: one agent routing to specialized agents? Where are gates and guardrails needed (user identification, queue management, handover to a human operator)? You design the structure before building it, across chat and voice, which follow the same logic.
- Define the standard, not just the solution. Your architectural choices become the team's default: playbooks, quality criteria, reusable patterns. What you solve once should not be reinvented on the next project.
- Control hallucinations, and turn it into a method. You define the boundaries of every agent: when it answers, when it retrieves information, when it must say "I don't know" instead of inventing a plausible answer. This is where an assistant's reliability is decided, and you establish this discipline, evaluation sets and regression tests included, for everyone.
- Translate business into behavior. You turn a business goal into concrete agent behavior: "cut queue times" or "never mention internal details to the user" become precise rules, flows and guardrails.
- Own the user experience. You design tone, messages and per-channel adaptation (voice, chat, messaging) from the point of view of the person living the interaction: perceived quality is part of the solution.
- Direct execution with AI. You use agentic coding tools (e.g. Claude Code) to build, and you read and validate what they produce. The craft is framing the problem so the machine executes it well, and knowing how to judge when the output is right and when it is not. And you teach the team to do the same.
- Ship to production and iterate. Solutions go live: you monitor performance and KPIs, manage edge cases, control the quality and reliability of the answers and iterate after go-live. The outcome is your responsibility, on your projects and on the team's.
- Grow people. You onboard and mentor your reports, set up 1:1s and growth plans, and take part in selecting the next hires. You define what "great" means in this craft and how it is measured.
- Guide the client through decisions. You are the technical point of reference in front of enterprise decision-makers: you help the client make the right choices, give the verdict on feasibility and scoping including in pre-sales, and defend the solution when timelines come under pressure.
- Coordinate the right people. You work closely with Product, DevOps, Customer Success and Technical Support to align, manage dependencies and unblock things fast. And you bring to Product, in a structured way, what the platform needs to become.
You will need:
- A track record with AI Agents in production. You have taken at least one agentic system live with real users and volumes, and you can tell its numbers: autonomous resolution rate, response times, perceived quality, cost per conversation.
- Agentic architectural judgment. You can design an agentic system end to end (frame it, decide where and how to use tools and context retrieval, design the flows and the guardrails) and explain the trade-offs of every choice, not just that it works.
- Reliability control. You can recognize and prevent hallucinations: define boundaries, understand when an agent should retrieve, answer, or admit it does not know. Answer quality is your design responsibility, and you can turn it into a repeatable method for the team.
- Technical leadership experience. You have already led engineers: mentoring, growth, technical decisions and priorities. Even informally, but with verifiable results.
- A functional, product/UX-oriented mindset. You design from the end user's point of view and bridge business and solution: you can translate a goal into concrete agent behavior.
- The ability to direct AI to build. You use agentic coding tools (e.g. Claude Code) daily and leverage AI to move faster: you frame the problem for the machine to execute it, and you read and judge the output well enough to be autonomous.
- Solid engineering foundations and advanced Python. You understand APIs, integrations and distributed systems well enough to design them and discuss them as a peer with Product and with the client's IT, and you read and write Python fluently: enough to build what you need, dig into what your tools produce, and go deep when the problem demands it.
- Production awareness. Comfortable with at least one cloud (GCP or AWS), with observability (logging, tracing, KPIs instrumented from day one) and with enough command of authentication, privacy and security, including GDPR and the EU AI Act, to design privacy-by-design solutions and be credible with European enterprise clients.
- Autonomy and communication. Structured problem solving on multi-stakeholder enterprise projects, with external dependencies and timelines not always in your hands. Excellent communication, written and spoken: native Italian and excellent English.
We will really like (but they are not required):
- Experience with conversational platforms / chatbots and with voice assistants (CCaaS).
- Having built a team from scratch, or defined delivery processes in a fast-growing scale-up.
- Experience with no-code/low-code automation platforms (e.g. n8n, Make, Zapier) and the ability to prototype workflows quickly.
- An ML/Data Science background: you understand the models, not just their APIs.
- Experience with enterprise clients and familiarity with project management and collaboration tools (e.g. Jira, Notion, Zendesk, Figma).
- Spanish and/or Portuguese, useful for working with our teams and clients in Iberia and Latin America.
- Tech talks, workshops, open source contributions: you know how to transfer what you know.
Our values:
At indigo.ai we prize Vision (curiosity and courage to challenge the status quo), Connection (empathy, candor, and trust), Responsibility (ownership, reliability, and follow-through), and Excellence (the habit of raising the bar and refining until it's right). If you naturally think ahead, build strong relationships, take accountability, and obsess over the quality of what you deliver, we're probably a great match.
What do we offer?
- A leadership role in a fast-growing European AI scale-up, backed by Azimut (€15M investment), where your work turns into products used by millions.
- A competitive salary in the range of €55–68k RAL, plus a €6k performance-based variable.
- A real mandate: you define how AI Agent delivery is done here, you do not inherit a process to execute.
- Concrete room to grow as we scale internationally: the team, the projects and the scope of this role are all expanding.
- A dedicated education budget and a structured Career Development Plan, in a culture that values transparency and continuous learning.
- Top-grade equipment (MacBook Pro and iPhone).
- A dynamic, people-first environment with great attention to well-being, and company retreats in stunning locations throughout the year.
Where is the job?
Milan, in a hybrid model. The team works together in the office: this is a role that lives on conversations with Product, Customer Success and clients, and being in Milan is part of the job. Our office is located at SPACES, Piazza Gae Aulenti 1/Torre B. We are scaling the team's capacity to support new projects in Italy and internationally.
How To Apply
We do not care how you got here: what counts is the demonstrated ability to design, make things work and grow people, not the academic path. So in your application, tell us two things:
- An assistant, an agent or a flow you designed yourself: the decisions you made, where it could fail, how you made it reliable, what numbers it produced in production.
- A person or a team you grew.
That tells us much more than any CV or title.
Это сохранённая копия объявления, опубликованного в другом месте. Вакансии снимают без предупреждения — перед откликом проверьте сайт работодателя. mentors.coach не является нанимающей стороной.