Senior ML Engineer

Afori; Barcelona; Presencial; EUR 90.000 - 150.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Relocation support

Beckham regime guidance

Relocation bonus

NIE & padrón support

Office in Barcelona

Flexible start date

Learning support

Equipment provided

Team culture

Descripción de la vacante

Afori, in Barcelona, is seeking an experienced ML Engineer to lead the learning and deployment of agent-based NLP models for insurance. You will design the learning flywheel, fine-tune models, and build cross-sell signals while upholding robust evaluation standards.

Join a fast-moving, equity-backed startup with relocation support to Spain, a flexible start date, and a path to meaningful ownership as you scale AI in broker workflows.

Formación

• Proven track record shipping ML systems in production.
• Experience designing evaluation methodologies (gold sets, benchmarks, abstention checks).
• Strong ML engineering fundamentals: training pipelines, serving infra, cost-aware inference.
• Comfort owning ambiguous problems end to end in a fast-moving environment.

Responsabilidades

• Learning Flywheel Architecture: design and own the continuous learning loop for Afori's agent swarm.
• Model Fine-Tuning & Distillation: manage fine-tuning and distillation for German insurance NLP tasks with evaluation gates.
• Cross-Sell Intelligence Models: build models to identify cross-sell opportunities from customer data.
• ML Infrastructure & Evaluation Standards: define evaluation methods and serving infra for production models.

Conocimientos

ML systems in production

Evaluation methodologies

ML engineering fundamentals

Ambiguity ownership

German language skills

Descripción del empleo

Help change one of the largest global industries: insurance brokerage.

About Us

Afori's vision is to ensure every human has the best cover before it's needed. Our founders previously built insurance unicorn wefox and are now building the AI layer for insurance brokers: agents that plug directly into brokers' daily workflows and take over the back-office work that eats more than 60% of their time. We already save brokers over 2.5 hours a day, freeing them up to do what actually grows the business: advising customers well.
The opportunity is massive: the global protection gap is €1.3 trillion, and insurance distribution still runs on manual processes. Sequoia recently mapped AI-native services opportunities (https://sequoiacap.com/article/services-the-new-software) and ranked insurance brokerage first, a $140-200B market. We raised a €4M pre-seed led by General Catalyst, backed by investors and operators including Mustafa Suleyman (CEO, Microsoft AI) and Alex Rinke (Co-CEO, Celonis)

The Role

You own how Afori's agents think. Not the interface and not the infrastructure, but the reasoning itself, and the economics of running it at scale across thousands of real insurance Cases in a regulated market.

Key Responsibilities:

• Learning Flywheel Architecture: Design and own the continuous learning loop for Afori's agent swarm: capturing agent outcomes, expert feedback, and case feedback to systematically improve agent performance over time.
• Model Fine-Tuning & Distillation: Own the fine-tuning and distillation pipeline for specialized German insurance NLP tasks: selecting, training, and evaluating open-weight and proprietary models per task, with rigorous eval gates before production rollout.
• Cross-Sell Intelligence Models: Build and own ML models that identify and score cross-sell opportunities across the customer book, translating APDB and account data into actionable revenue signals for the broker.
• ML Infrastructure & Evaluation Standards: Establish and maintain the evaluation methodology (gold sets, benchmarks, abstention/faithfulness checks) and the serving infrastructure that all Afori ML models must meet before deployment.

What We’re Looking For

• Strong track record building and shipping ML systems in production: fine-tuning, distillation, and evaluation of both open-weight and proprietary LLMs.
• Hands-on experience designing evaluation methodologies: gold sets, benchmarks, abstention and faithfulness checks, and eval gates that decide what reaches production.
• Solid ML engineering fundamentals: training pipelines, serving infrastructure, and cost-aware inference at scale.
• Comfort owning ambiguous problems end to end in a fast-moving early-stage environment. You are a finisher, not just a starter.
• Bonus: German language skills, NLP experience in regulated domains (insurance, finance, legal, health), or experience with agentic systems and continuous learning loops.

Compensation & Benefits

We let you choose your own balance between cash certainty and ownership: a higher-cash or a higher-equity package, with the equity-weighted option carrying a deliberate premium in total value.

• Base salary: Competitive
• Team bonus: 10% of base at 100% team goal achievement, paid half in cash and half in equity.
• Equity: A meaningful founding-level option grant in a4i Inc. (Delaware C-Corp TopCo), vesting over 4 years with a 1-year cliff, then monthly. Options convert to common shares on exercise, making you a real shareholder on the cap table.
• Annual reviews: Compensation and options are rediscussed after 12 months at the latest. We grant options/shares annually, so the best performers continuously get more.

Pathways to additional equity

• Annual company ranking: Top performers receive additional equity at the annual compensation review.
• Team goal achievement: When teams exceed objectives, you choose between a cash bonus and additional equity.
• Promotion: Grants aligned with expanded responsibility.
• Direct investment: You may invest personal funds into a4i Inc. Shares are issued immediately with no vesting.

Relocation & Benefits

Relocating to Barcelona? We handle it. If you move your tax residency to Spain, you may qualify for the Beckham regime (flat 24% income tax up to €600,000 for up to six years). Afori prepares, files and pays for the application, covers the gestoria review, provides a relocation bonus, and supports your NIE, Social Security and padrón registration.

• Work policy: Office-first culture in Barcelona, with flexibility.
• Paid time off: 24 days annual leave plus public holidays, rising by one day per year of service to a maximum of 30.
• Learning support: We fund courses, conferences and certifications that help you develop, approved case by case.
• Equipment: Whatever you need to do the work properly.
• Team culture: Regular team events and genuine Dream Team collaboration.
• Flexible start date: We accommodate your notice period and, if needed, your move.

The Process

Speed matters to us. With the right person, we can run the whole process in under two weeks.

• Intro call with Felix, founding team member
• Interview with Sergi, our Co-Founder & CTO
• Interview with Fabian, our CEO & Co-Founder
• One-day trial with us in the Barcelona office