Forward Deployed AI Engineer

About Vectrix

Vectrix builds supervised AI agents for transport and logistics operations.

Today, our platform is best known for transforming incoming emails, documents and spreadsheets into structured transport orders. We are now expanding beyond order entry toward modular AI agents that support operational work across customer service, planning, warehousing, documents, billing, claims, customs and fleet operations.

These agents work across the systems logistics teams already use, including mailboxes, TMS platforms, ERPs and spreadsheets. Clear work can be completed automatically. Ambiguous work is prepared for human review, so people remain in control of decisions, exceptions and customer relationships.

Our ambition is to help transport organizations increase capacity without adding the same amount of manual coordination.

Why now

Order automation is our starting point. The next stage is discovering and deploying AI agents across the wider logistics organization. We need people who can find where automation creates genuine value and turn those opportunities into reliable Production workflows.

The role

As a Forward Deployed AI Engineer, you will own the technical side of customer deployments.

You will build the technical foundation and customer solutions that allow modular AI agents to operate safely across mailboxes, transport-management systems, ERPs, documents and other operational tools.

You will move between customer context and the codebase: investigating failures, building integrations, improving supervision and observability, and turning customer-specific solutions into reusable product capabilities. This is not an engineering role isolated from users.

Help build the AI operating layer for transport and logistics.

What you will do

• Work directly with customers to understand their workflows, systems, data and operational constraints.
• Translate operational requirements into technical designs and production implementations.
• Build integrations with TMS, ERP, CRM, email, document and other enterprise systems.
• Develop prototypes that validate technical feasibility and customer value.
• Own technical delivery from discovery through UAT, go-live and early Production.
• Design modular agent workflows using LLMs, tools, prompts, rules and structured outputs.
• Build safe human-supervision, approval and exception-handling flows.
• Diagnose failures across documents, data, prompts, models, integrations and application code.
• Improve agent reliability through tests, evaluations, observability and feedback loops.
• Build internal tools that make deployments faster, safer and more consistent.
• Turn recurring customer requirements into reusable agents and platform capabilities.

What success looks like

• Customer workflows move from prototype to stable Production systems.
• Integrations and technical blockers are resolved end to end.
• Deployment learning becomes scalable product capability, tooling and observability.

Own the outcome. Surface risks early. Solve the underlying problem.

Who we are looking for

You have at least three years of experience building and shipping production software, or an exceptional equivalent track record. Relevant backgrounds include full-stack or backend engineering, integration engineering, hands-on solutions engineering, data or ML engineering, technical consulting and early-stage startup engineering.

You should be comfortable switching between deep technical work and direct customer collaboration.

• Strong professional experience in Python or TypeScript and the ability to become productive in the other.
• Experience designing APIs and integrating external systems.
• A solid understanding of application architecture, data modelling and databases.
• Experience diagnosing complex behaviour across services and data sources.
• Evidence that you can ship maintainable, tested production software, not only prototypes.
• Excellent communication with technical and non-technical stakeholders.
• The confidence and curiosity to learn unfamiliar customer domains quickly.
• High ownership from customer problem through technical resolution.
• Professional fluency in Dutch and English.

Experience with LLM applications, AI agents, prompt engineering or evaluation is strongly preferred, but learning speed matters more than years in a recently emerging field. Svelte, Convex, FastAPI, SQL, document processing, observability and logistics systems are additional advantages.

A BSc or MSc in computer science, engineering or a related discipline is preferred. Equivalent practical achievement is equally welcome.

Why Vectrix

Join Vectrix while we expand from automating transport-order entry into building the AI operating layer for logistics. You will work directly with the founders, customers and engineering team, with room to experiment, learn and take ownership.

Our office is in central Antwerp, a five-minute walk from Antwerp Central Station and located within a thriving technology community. We work together from the office four days per week, with occasional customer visits.

We value high ownership, customer empathy, analytical rigour, warmth and speed. We take responsibility for outcomes, surface problems early and solve them without blame-shifting.

We offer competitive compensation. We expect this role to participate meaningfully in Vectrix’s ESOP: the people who create long-term value should share in the outcome.

Candidates must already be authorized to work in Belgium. We do not provide visa sponsorship.

Hiring process