- Could you introduce yourself and tell us about you? Your current / recent jobs and employers? Previous companies and jobs?
He has 1.5+ years of industry experience and graduated from the XXXX with a mathematics degree on the computer science track, focusing on machine learning and deep learning. A few months before graduation, he became involved in a friend’s startup, XXXX, starting as an adviser and later taking on coding work. The team explored different ideas, including a financial Q&A Discord server, a website version of the tool, and later a small B2B effort involving Amazon data scraping. The company eventually dissolved due to the co-founders moving on, and during this period he was also working part-time, contributing around 10–15 hours on some weekends. He later joined XXXX , where one of his key achievements was developing and shipping an AI chatbot for question answering. He was the only one handling AI work, so he designed the backend architecture, created API classes, integrated Google Vertex for function calling, and built a FastAPI service for session management. Earlier at XXXX he also built a retrieval system inspired by research from XXXX University and created a middleware class that integrated AI into Discord for handling chat sessions. These projects highlight his experience in backend development, AI integration, and designing practical solutions for both startup and product environments.
- What skill(s) / experience would you self-describe as strongest or specialist in?
His strongest skills are in machine learning, deep learning, and backend development. He is highly proficient in PyTorch and scikit-learn, with hands-on experience building custom model layers, retrieval systems, and scikit-learn–compatible classes.
- Do you have the lists of key criteria’s according to the clients’ requirements?
He has solid experience with machine learning libraries, mainly PyTorch and scikit-learn. With PyTorch, he practiced by implementing different model layers such as group query attention, multi-query attention, and LoRA. He also worked on several projects that strengthened his understanding of model structures. For scikit-learn, he completed multiple academic tasks and even built a scikit-learn–compatible class for a clustering algorithm when the original implementation didn’t meet the framework’s standards. This project became a production-quality clustering model in his portfolio. He has some experience with Keras as well, and he commonly works with Python, also, not much with C++, and SQL but if requires him to work with them he have enough practice with them to pick them up quickly.
