Data Scientist

About the company

Forto is working to transform global logistics through AI and digital supply chains. The goal is to make international trade simpler, more transparent, smoother, and more sustainable by improving the systems that support freight and logistics operations.

Role overview

In this position, you will own machine learning systems that convert unstructured logistics information into structured outputs ready for production use. You will work closely with the Engineering Manager across three main areas: document extraction (FlashDoc), vocabulary mapping, and classical machine learning. A major focus will be keeping current production systems reliable while building robust evaluation pipelines. At the same time, you will be expected to push accuracy improvements by exploring stronger methods and newer technologies whenever the opportunity exists.

In addition to document automation, the wider roadmap includes more traditional data science problems such as demand forecasting, churn prediction, route optimisation, and predictive analytics for logistics operations. The role therefore combines ML engineering depth with classical data science capability.

Key responsibilities

• Develop, deploy, and support production-grade ML workflows for document extraction, classification, and data enrichment.
• Create evaluation systems and feedback-driven optimisation loops to improve extraction quality in a structured way.
• Use human-in-the-loop data to train internal models that help move extraction from assisted to fully automated.
• Build and maintain semantic matching models that map free text into standard TMS terminology for items such as ports, terminals, container categories, legal entities, and line items.
• Increase system resilience through redesign, testing, observability, monitoring, and alerting for non-deterministic ML behaviour.
• Assess and adopt advanced approaches such as new model architectures, fine-tuning methods, and novel evaluation techniques when incremental improvements are no longer enough.
• Work with Product Managers to identify practical data science opportunities, uncover needs from the data, and contribute to product direction.
• Partner with Engineering teams on integration, infrastructure, and API design so downstream systems can reliably consume DS outputs.
• Set clear expectations with stakeholders, communicate feasibility and delivery timelines, surface risks early, and help prioritise work across teams.

Requirements

• At least 2 years of professional experience in data science or machine learning engineering.
• Proven ability to design, ship, and maintain ML systems in production, including pipeline architecture, monitoring, reliability, and handling non-deterministic outputs at scale.
• Ability to quickly ramp up on new tools, technologies, and problem domains.
• Comfort using agentic coding tools.
• Strong Python skills.
• Practical experience with LLMs, including prompting, fine-tuning, and evaluation, along with awareness of their production limitations.
• Solid grounding in classical data science and statistics, including regression, classification, time series analysis, data leakage, experimental design, and hypothesis testing.
• Strong analytical thinking and problem-solving ability, with the capacity to work independently on ambiguous, research-heavy problems.
• Ability to recognise when existing methods are not enough and propose more disruptive alternatives rather than only making incremental improvements.
• Strong stakeholder management skills, including proactive problem identification, expectation setting, and prioritisation trade-off negotiation.
• Ability to commit to a final direction after healthy debate, even if it is not the preferred option.

Culture and application note

Forto encourages applicants who do not match every requirement to still consider applying, especially if they are motivated by the role and the company’s values. The organisation values diverse perspectives, clear communication, efficient systems, and a collaborative way of working that balances speed with care.

Work location

This is an onsite position based in Berlin, Germany.