We use cookies to run the site and count visits. "Got it" also allows session recording.
KKI•PIE Group
Bad Homburg, Germany4 months ago
Grade
Senior
Work format
Hybrid
Employment
Full-time
English level
B2, upper-intermediate
An email or a Telegram message to the recruiter
Context: KI•PIE Group has been providing price data and analysis for the European plastics and chemical industry for over 50 years. On October 12, 2026, we launched pricenxt.com, our SaaS platform for pricing and market analysis. You will develop the AI layer on top of it: using proprietary market data that no one else has, and for paying users who will provide feedback from day one.
As a (Senior) LLM Application Engineer, you will be responsible for our LLM platform from concept to implementation. In the first few months, the focus will be on AI features for pricenxt customers. This will be followed by internal tools to accelerate the work of our editorial and sales departments.
Real proprietary data and real users, not a test environment. A recently launched product where your work will be directly visible. A small international team, flat hierarchy, short decision-making paths.
You will build the LLM platform layer: backend for search, embedding pipeline, and clean APIs. You will implement AI features in production: Conversational Analyst for our price data, semantic search, and intelligent alert flow. You will integrate foundational models (Anthropic, OpenAI, Mistral, etc.) and make decisions about model selection based on quality, cost, and latency. You will build agent architectures and MCP interfaces through which corporate clients can receive our data directly into their systems. You will participate in decisions about which use cases will be implemented in production and measure their effectiveness: evaluation, ensuring ground truth, monitoring.
Several years of experience as a Software Engineer, of which at least one LLM function that you have implemented in production yourself and which is used by real users. Experience with RAG in production: Vector DB, Retrieval, Chunking, Evaluation, and Grounding. Proficiency in backend development with Python and/or TypeScript/Node: Services, Async, API Design, Data Modeling. You are familiar with Function Calling, Tool Use, and Prompt/Context Engineering in practice, not just from textbooks. You work independently, make technical decisions, and can explain them to non-technical specialists. Very good knowledge of German and English.
MCP (Model Context Protocol) or frameworks/SDKs for agents. React/TypeScript, so you can independently create user interfaces for your features. Data Pipelines for structured and semi-structured sources. Experience with market data, time series, or B2B SaaS.
Permanent employment is preferred. For a suitable candidate, a transition to freelance or a contract for 6-12 months is also possible.
Attractive salary plus annual bonus Hybrid work with an office in Bad Homburg and flexible hours. Budget for models, tools, and conferences. JobRad, JobAuto, and JobTicket. Company pension plan (bAV) and supplementary corporate health insurance (bKV). Corporate Benefits, as well as at least one team event per quarter.
Context: KI•PIE Group has been providing price data and analysis for the European plastics and chemical industry for over 50 years. On October 12, 2026, we launched pricenxt.com, our SaaS platform for pricing and market analysis. You will develop the AI layer on top of it: using proprietary market data that no one else has, and for paying users who will provide feedback from day one.
As a (Senior) LLM Application Engineer, you will be responsible for our LLM platform from concept to implementation. In the first few months, the focus will be on AI features for pricenxt customers. This will be followed by internal tools to accelerate the work of our editorial and sales departments.
You will build the LLM platform layer: backend for search, embedding pipeline, and clean APIs. You will implement AI features in production: Conversational Analyst for our price data, semantic search, and intelligent alert flow. You will integrate foundational models (Anthropic, OpenAI, Mistral, etc.) and make decisions about model selection based on quality, cost, and latency. You will build agent architectures and MCP interfaces through which corporate clients can receive our data directly into their systems. You will participate in decisions about which use cases will be implemented in production and measure their effectiveness: evaluation, ensuring ground truth, monitoring.
What you bring: Several years of experience as a Software Engineer, of which at least one LLM function that you have implemented in production yourself and which is used by real users. Experience with RAG in production: Vector DB, Retrieval, Chunking, Evaluation, and Grounding. Proficiency in backend development with Python and/or TypeScript/Node: Services, Async, API Design, Data Modeling. You are familiar with Function Calling, Tool Use, and Prompt/Context Engineering in practice, not just from textbooks. You work independently, make technical decisions, and can explain them to non-technical specialists. Very good knowledge of German and English.
What would be an asset: MCP (Model Context Protocol) or frameworks/SDKs for agents. React/TypeScript, so you can independently create user interfaces for your features. Data Pipelines for structured and semi-structured sources. Experience with market data, time series, or B2B SaaS. Permanent employment is preferred. For a suitable candidate, a transition to freelance or a contract for 6-12 months is also possible.
Attractive salary plus annual bonus Real proprietary data and real users, not a test environment. A recently launched product where your work will be directly visible. A small international team, flat hierarchy, short decision-making paths. Hybrid work with an office in Bad Homburg and flexible hours. Budget for models, tools, and conferences. JobRad, JobAuto, and JobTicket. Company pension plan (bAV) and supplementary corporate health insurance (bKV). Corporate Benefits, as well as at least one team event per quarter.
We compare skills, grade, work format and salary
Most often 73,100 to 100,000
