Making LLMs, deep learning and data analysis useful for you
A registered swiss business: CHE-319.390.704
Data analysis on massive datasets, tailored to your application. Data really is key for any analysis and eventual deep learning model development. Let's focus on collecting, filtering and organizing data on a massive scale. This involves using object storage offerings from cloud providers to scale up and down the data stored effectively at reasonable costs. Recent LLMs advancements in efficiency also allow us to filter and organizing data at extreme accuracy, by having the machine "understand" the data. While still a rather expensive operation, this one-time cost of intelligence can really elevate the results.
Fast prototyping of RAG and data management systems to apply the latest advancements in AI to your specific niche. By working together, we can boost accuracy and usefulness of the best models in the world, and ensure reliability (practically zero hallucinations).
Staying current across the entire technology stack is key. From identifying the best current hardware for price and performance, to selecting a Python version that works well with specific libraries and cuda drivers, right down to algorithmic optimizations, I maintain the full stack of practical and theoretical knowledge needed to deliver top results efficiently, considering cost, ease of deployment and speed of execution.
Rapid prototyping of user-friendly applications. Expertise in Flutter for cross-platform mobile apps (Android & iOS) and web development using React and Flask, typically. Proficient in integrating databases, including SQLite for quick prototypes, and experienced in deploying robust production applications on Linux servers.
Consumer-oriented website allowing anyone to use the current best models, without paying a subscription.
Performed market and sentiment analysis for Keyvian, a startup incubated at Blaze (EPFL's startup incubator). The project involved scraping massive amounts of web data from 2013 to 2023 and comparing it to price movements in various crypto assets during that period.
Sacha Liechti
BSc Computer Science, EPFL
MSc Computer Science Candidate, ETH Zurich