devjobs.de

Senior Solution Engineer

deepset GmbH · Berlin | Deutschland · Hybrid, Onsite

Gehalt auf Anfrage

Gefunden am 19.05.2026

Match

84%

Fit in Skills, Kultur und Entwicklungspfad.

Beschreibung

Job Zusammenfassung In dieser Rolle entwickelst du praktikable NLP-Lösungen für Kunden, evaluierst und optimierst Modelle, gestaltest Annotation-Workflows und führst Workshops, um dein Wissen über Haystack und NLP zu Job Zusammenfassung In dieser Rolle entwickelst du praktikable NLP-Lösungen für Kunden, evaluierst und optimierst Modelle, gestaltest Annotation-Workflows und führst Workshops, um dein Wissen über Haystack und NLP zu teilen. Deine Rolle im Team As a Solution Engineer at deepset, you will apply cutting-edge NLP techniques to real-world enterprise problems. You will develop deep expertise across the full spectrum of NLP methods within Haystack including dense retrieval, Question Answering, generative QA, Summarization, RAG pipelines, and more and use deepset Cloud, our SaaS platform, to streamline workflows and transform customer data into actionable intelligence. You will collaborate closely with our open-source and product teams to shape the best solutions for each engagement. This role sits at the intersection of professional services consultant, senior AI/NLP engineer, and trusted technical advisor. Exemplary engagement: You will partner with data scientists from a leading aerospace corporation to enhance search functionality across pilot manuals. This means understanding their data and domain, evaluating methods (DPR, QA, TableQA) against their requirements, designing annotation workflows for expert labellers, continuously analysing incoming labels, training and iterating models, and driving step-by-step performance improvements until production targets are met. Show customers how to search for and extract information using Haystack and deepset Cloud, growing NLP awareness within their organisation. Understand each customer's domain data, use case, needs, and processes to craft the optimal solution and drive improvements to Haystack/deepset Cloud. Be the outward face to the customer: communicate to define goals and timelines, and report results throughout the engagement. Rapidly prototype cutting-edge NLP methods - dense retrieval (DPR), Question Answering, generative QA, TableQA, Summarization - and evaluate them against real customer data. Lead end-to-end technical delivery of AI solutions using Haystack, PEFT, Hugging Face Transformers, and deepset Cloud. Design annotation workflows for expert labellers and continuously analyse incoming labels to drive model improvement. Train and fine-tune large language models through large-scale data curation and optimisation to meet production performance targets. Contribute NLP/RAG-related improvements upstream to Haystack and other open-source libraries. Partner directly with customer data scientists and engineering teams, guiding them from prototype through to production. Run technical workshops and enablement sessions; help customers build internal capabilities around Haystack and NLP best practices. Act as the internal voice of the customer, feeding domain knowledge and product feedback to our open-source and product teams. Track performance metrics and communicate progress clearly to both technical and non-technical stakeholders. Define and pursue the skills needed to grow into an expert or team lead role as deepset scales. Contribute to deepset's technical blog, conference talks, and Haystack community presence. Work with Account Executives and the Product team to translate customer insights into roadmap decisions. Unsere Erwartungen an dich Ausbildung University degree in Computer Science or a comparable qualification. Qualifikationen Fluency in German (C1/C2 or native) and strong professional English - you will operate in both languages daily. Proficiency in Python and strong working knowledge of ML/DL/NLP methods and recent model architectures (Transformers, LLMs, retrieval models). Excellent communication skills with both technical (data scientist, ML engineer) and non-technical (business, C-suite) stakeholders. An intense desire to learn and a track record of rapidly acquiring new skills. Familiarity with dense retrieval (DPR), Question Answering, generative QA, TableQA, and Summarization methods. Understanding of RAG architectures and vector databases (e.g. Weaviate, Pinecone, OpenSearch). Contributions to open-source NLP libraries or a public technical profile (blog, talks, GitHub). Aspiration to grow into a senior expert or team lead position as the Solutions team scales. Erfahrung 5+ years of industry experience applying data science or NLP methods to real-world data. Hands-on experience with Haystack, Hugging Face Transformers, PEFT, or comparable NLP frameworks. Experience managing customer-facing projects: defining scope, setting timelines, communicating results. Data science consulting experience and/or hands-on experience with data annotation workflows. Experience with deepset Cloud or other enterprise NLP/search platforms. Experience with model fine-tuning via PEFT (LoRA, QLoRA) and large-scale data curation. Unser Angebot Competitive salary + equity (ESOP). We benchmark against above European AI companies. Remote-first culture with optional access to Munich and Berlin offices. Flexible working hours. Work on the frontier of enterprise AI. Your solutions will go into production at household-name companies. Learning budget, conference attendance, access to the best minds in open-source NLP and LLMs. Transparent, low-ego team. English as the working language, diverse and international colleagues. 30 days PTO, company offsites, home office setup budget, and more. Benefits Work-Life-Integration 🏠 Home Office Themen mit denen du dich im Job beschäftigst Machine Learning AI Metadaten Level: Senior Job Feld: Data, Application Anstellung: Vollzeit Vertragsart: Unbefristetes Dienstverhältnis Arbeitsmodell: Hybrid, Onsite Unternehmenstyp: Startup Branche: Internet, IT, Telekom Ort: Berlin | Deutschland

Tech Stack

Python

Warum passt du zu dieser Stelle?

Fit technisch: Stark auf Backend und API-Architektur.

Gaps: Fehlende Tool-Erfahrung in 1-2 Schlüsselbereichen.

Success-Wahrscheinlichkeit: Hoch bei schneller Einarbeitung.