Berlin, Berlin, Germany
Kineo assists companies at different stages of their journey towards becoming data and AI-driven organizations, striking a delicate balance between data/ML expertise and business focus. Thereby, Kineo takes a comprehensive and unbiased approach to help companies explore the potential of cutting-edge machine learning technologies, offering professional guidance that is applicable across various industries and domains. During my tenure at Kineo, I had the opportunity to work on several impactful use-cases, including: Quality Assurance based on Computer Vision: Quality assurance is a critical concern for our primary customer segment, which comprises manufacturing companies, as it directly impacts their core business operations. Addressing this challenge involves navigating a triad of obstacles: the demand for rapid inference speed, a scarcity of labeled images, and the imperative of achieving high target accuracy. By leveraging a pretrained Vision Transformer model we were able to surpass existing solutions in terms of accuracy, while requiring only a minimal amount of labeled training data. This strategic approach swiftly generated tangible business value. Moreover, deploying the model on the edge with GPU support enabled us to achieve remarkable inference speeds. Document Summarization based on Large Language Models: Many of our clients have shown a keen interest in applications related to NLP. With the release of ChatGPT, this interest has significantly intensified due to the promising capabilities of LLMs. Despite facing stringent data privacy regulations, we have successfully incorporated Prompt Engineering solutions using Retrieval Augmented Generation (RAG) methods as well as fine-tuned open-source models using Parameter Efficient Fine Tuning (PEFT).
I played a pivotal role in co-creating a sophisticated distribution model specifically tailored to tackle the supply chain challenges faced by a leading heating technology manufacturer. This solution directly addressed the repercussions of global shortage issues that were even amplified by vast demands in the market.