Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm Sneha Banerjee, an AI Engineer and researcher based in Leonberg, Germany, focused on building explainable AI systems and interactive ML applications. I design Retrieval-Augmented Generation pipelines, semantic search, and multi-turn memory for QA and recommendation tasks, using tools like FAISS, LangChain, OpenAI GPT-3.5, PyTorch, and Streamlit. Currently, I work on self-initiated projects on explainable document QA and semantic book recommendations, and I have hands-on experience migrating DL pipelines from TensorFlow to PyTorch, building browser-based 3D viewers, and delivering user-friendly GUIs. I enjoy collaborating across teams to create transparent, practical AI solutions.…I'm Sneha Banerjee, an AI Engineer and researcher based in Leonberg, Germany, focused on building explainable AI systems and interactive ML applications. I design Retrieval-Augmented Generation pipelines, semantic search, and multi-turn memory for QA and recommendation tasks, using tools like FAISS, LangChain, OpenAI GPT-3.5, PyTorch, and Streamlit. Currently, I work on self-initiated projects on explainable document QA and semantic book recommendations, and I have hands-on experience migrating DL pipelines from TensorFlow to PyTorch, building browser-based 3D viewers, and delivering user-friendly GUIs. I enjoy collaborating across teams to create transparent, practical AI solutions.

Sneha Banerjee

AI Engineer, Data Scientist, Developer, +2





I'm Sneha Banerjee, an AI Engineer and researcher based in Leonberg, Germany, focused on building explainable AI systems and interactive ML applications. I design Retrieval-Augmented Generation pipelines, semantic search, and multi-turn memory for QA and recommendation tasks, using tools like FAISS, LangChain, OpenAI GPT-3.5, PyTorch, and Streamlit. Currently, I work on self-initiated projects on explainable document QA and semantic book recommendations, and I have hands-on experience migrating DL pipelines from TensorFlow to PyTorch, building browser-based 3D viewers, and delivering user-friendly GUIs. I enjoy collaborating across teams to create transparent, practical AI solutions.…I'm Sneha Banerjee, an AI Engineer and researcher based in Leonberg, Germany, focused on building explainable AI systems and interactive ML applications. I design Retrieval-Augmented Generation pipelines, semantic search, and multi-turn memory for QA and recommendation tasks, using tools like FAISS, LangChain, OpenAI GPT-3.5, PyTorch, and Streamlit. Currently, I work on self-initiated projects on explainable document QA and semantic book recommendations, and I have hands-on experience migrating DL pipelines from TensorFlow to PyTorch, building browser-based 3D viewers, and delivering user-friendly GUIs. I enjoy collaborating across teams to create transparent, practical AI solutions.

Available to hire

I’m Sneha Banerjee, an AI Engineer and researcher based in Leonberg, Germany, focused on building explainable AI systems and interactive ML applications. I design Retrieval-Augmented Generation pipelines, semantic search, and multi-turn memory for QA and recommendation tasks, using tools like FAISS, LangChain, OpenAI GPT-3.5, PyTorch, and Streamlit.

Currently, I work on self-initiated projects on explainable document QA and semantic book recommendations, and I have hands-on experience migrating DL pipelines from TensorFlow to PyTorch, building browser-based 3D viewers, and delivering user-friendly GUIs. I enjoy collaborating across teams to create transparent, practical AI solutions.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Language

English

Fluent

German

Intermediate

Bengali

Fluent

Hindi

Advanced

Work Experience

Explainable Document QA Assistant with LLMs (Self-initiated Project) at Self-Initiated Project

August 1, 2025 - Present

Designed and implemented a Retrieval-Augmented Generation (RAG) pipeline to answer queries from a custom PDF knowledge base. Processed documents with FAISS and HuggingFace models, built an LLM-powered QA system using LangChain and OpenAI GPT-3.5, and integrated a Streamlit UI for interactive querying. Implemented evaluation pipeline measuring faithfulness, context precision/recall, and response latency. Enabled multi-turn conversational memory, similarity-score visualization, and retrieved-context explainability for better transparency.

Semantic Book Recommender with LLMs (Self-initiated Project) at Self-Initiated Project

July 1, 2025 - Present

Developed an interactive book recommendation system that provides content- and emotion-based suggestions. Demonstrated integration of OpenAI embeddings with local vector storage (ChromaDB) for efficient semantic search. Built a web-based interface using Gradio to input queries and filter results by category and tone (e.g., Happy, Sad, Angry), displaying book covers and concise descriptions dynamically.

Research Assistant at Fraunhofer IPA

June 1, 2024 - March 1, 2025

Developed a browser-based interactive 3D model viewer using Three.js, enabling loading, viewing, and interaction with 3D assets. Built a fully responsive GUI to demonstrate the capabilities of the bin-picking cell and interact with the 3D scene. Designed and styled the GUI with HTML5, CSS3, and React, using Vite for fast development. Accessed 3D model files and metadata from a backend REST API, handling error states and asynchronous loading. Migrated an existing deep learning library from TensorFlow to PyTorch, ensuring functional parity and improved readability. Optimized performance with PyTorch DataLoader, GPU acceleration, and memory management.

Research Assistant (Master’s thesis) at Fraunhofer IPA

November 1, 2023 - May 1, 2024

Developed autoencoder-based ML models to create a source library from large datasets using PyTorch and PCL, used in a transfer learning pipeline. Implemented and evaluated clustering algorithms to select representative members of the dataset. Extensive data preprocessing and dimensionality reduction for efficient handling of large-scale point cloud datasets.

Research Assistant at Fraunhofer IAIS

November 1, 2022 - October 1, 2023

Researched and implemented computer vision models for image classification, focusing on transformer-based architectures and deep CNNs in safety-critical perception tasks. Developed interactive and explainable classification systems using Grad-CAM and Shapley Value–based methods (ViT Shapley) to analyze concept and region relevance in autonomous driving scenarios. Proficiently used Git for version control, and maintained CI/CD pipelines with GitLab CI to automate testing, builds, and deployment.