AI Integration Engineer

Job Description Avanquest is the software publishing subsidiary of Claranova, a French group of innovative technological solutions, established in many countries, with revenues exceeding $500 million. At Avanquest, we create reliable and intuitive software solutions that satisfy our customers through 3 key product types: PDF software, photo editing software (Inpixio), and utility software to improve PC performance and security. Mainly offered as Software as a Service (SaaS), our products are sold in more than 150 countries and are translated into more than 15 languages worldwide. Avanquest is: a multicultural collaborative group; an ambitious company in rapid development; a place where each Talent has its place to contribute its know-how and have a real impact on large-scale projects; an environment where productive initiatives are always recognized at their true value. Visit our website : www.avanquestgroup.com Job details : We're building AI deeply into our product stack - from open-source model serving to the interfaces our users interact with every day . As an AI Integration Engineer, you'll own three interconnected layers : a model- serving REST API, AI- powered web app features , and the orchestration backbone of our document processing platform. This is a high- ownership , full-stack role for an engineer who wants to ship real AI systems , not just call third -party APIs. Model Serving & REST API Deploy open-source LLMs and expose them through a versioned , production-grade API. Web App Integration Embed AI capabilities into our web interfaces - streaming UX, real-time features , and AI- powered components. Document Processing Orchestration Build the processing pipelines that extract , classify , transform , and act on documents at scale . Key Responsibilities : Model Serving & API Evaluate and d eploy open-source LLMs via vLLM , Ollama , or TGI on GPU infrastructure ; Design and maintain a REST API with auth , rate limiting , versioning, and API doc umentation ; Optimize inference : quantization , batching , and concurrency tuning for latency / cost trade-offs ; Own observability - latency , throughput , GPU utilization , and cost -per-call dashboards . Web App Integration Integrate AI features into our web app s and document processing platforms ; Build AI- powered UI components: chat interfaces, inline suggestions, smart search , and summarization widgets ; Collaborate with frontend engineers on UX patterns specific to async and streaming AI responses ; Manage API keys, usage metering , and per-user AI feature flags on the client layer . Document Orchestration Design multi- step orchestration pipelines for document ingestion, parsing , classification, and extraction ; Integrate OCR, layout detection , and chunking strategies for PDFs , Word docs, spreadsheets , and scanned images ; Build RAG pipelines with hybrid retrieval (dense + sparse ) over document corpora ; Design and optimize prompts for extraction, summarization, classification, and conversational AI workflows; Implement task queues and async workers to handle high-volume document processing reliably . Platform & Collaboration Define evals and quality benchmarks for extraction accuracy , classification, and generation quality ; Write technical docs, so the platform is maintainable by the whole team and accessible for technology partners ; Partner with product , data science, and DevOps / MLOps to ship AI features reliably and at scale ; Stay current on the open-source model ecosystem and advise on model selection and upgrades. Requirements: 3+ years of software engineering experience spanning backend, API design, and web technologies ; Proven experience deploying or integrating LLMs in production - open-source model experience strongly preferred ; Hands-on knowledge of REST API design principles and at least one API framework ; Experience building or integrating AI features in a web app , including streaming/async response handling ; Familiarity with document processing workflows - parsing, OCR, extraction, or classification at scale ; Comfort designing multi-step orchestration pipelines with task queues, retries, and failure handling ; Strong communication skills; able to reason about trade-offs across latency, cost, and accuracy with stakeholders. What we offer: A fair compensation for your value; Bonus program, paid vacation leave and competitive corporate benefits; We are committed to maximizing your potential and ensuring your professional development. Interaction with local and international teams; A friendly and collaborative work environment, where authenticity and well-being are a priority.