Nenad Aksentijević

Supporting clients in defining and implementing generative AI use cases (customer feedback analysis, text documents) Leading technical discussions to improve user experience and guide AI decisions Designing solutions based on: OCR & Document AI for information extraction LLMs via APIs for question-answering and generation Multi-agent RAG with LangChain, LangGraph Deployment and industrialization (Google Cloud Platform, AWS...) Full-stack development and product integration (Python, FastAPI, Streamlit) Building AI POCs and demonstrators for clients Setting up low-code workflows with n8n to orchestrate AI use cases (OCR, LLM, RAG)

– Work in an agile environment on the application using techniques (NLP) and to optimize processes through feedback texts (for example, surveys or questionnaires). – Main task: collaborate with clients to improve the user experience, optimize and add new features. – Secondary task: creation of - with LangChain: – creation of an agent that uses RAG to retrieve data requested by the user from a VectorDB – creation of an agent that uses SQL queries to retrieve data requested by the user from the tabular data – creation of a user interface (like ChatGPT) to facilitate the use of the tool for clients

– Automating deployment in a cluster using LangChain: – for LLM (GPT-4) – Retrieval Augmented Generation (RAG) – Creation of tools that generate Kubernetes configurations in YAML files and automatically deploys the files in the cluster – Deployment of a secret analysis automation application for source code directories in the cluster using Docker, Kubernetes and Kyma Runtime – Securing access with Nginx authentication and OAuth2 – GitHub CI/CD pipeline