About

I started working with NLP in 2018 to find new methods of NLP that can have massive potential for the education ecosystem. After learning the baseline of NLP features, I created an exparimental API to test semantic search on nonfictions. This led me to understand that vector search heavily relies on text processing, which let me develop a few key principles for book data storing that improved search and semantic accuracy.

Fast-forward to the launch of GPT-3.5 API. I've started researching and evaluating the ability of generative models to accurately cite external sources. Through this research, I've participated in several studies that identified the power of combining semantic search with generative models, later formalized as the RAG architecture.

I continue to research, evaluate, and benchmark new models of LLMs for accurately citing external sources and develop testing methods to assess Context Length thresholds effectively. Additionally, with my background and love for philosophy and literature, I became susceptible to AI hallucinations, which in most cases look more like over-extrapolating the original source into digestible forms of commentary. In doing so, I've acquired a strong passion for the 'Red Team,' thinking of pulling out more and more cases of unfound needles in the haystack.

While the building blocks of generative AI models are improving every day, I believe that the key to successful integration is all about creating the right user experience to boost LLMs and make them more reliable.

For me, Web Design is first of all a thinking tool. A Clear Design is the first visual validation one can do with any digital product. Working as an html developer, I became more interested in design. After a year of training, Starting on my own project, I've designed my social reading app, from idea to production. After taking part in other digital projects, I started focusing on testing new ideas textual knowledge.

My design style is functional and independent. I believe that good design can be purified to minimal, where never save on data flowing with accurate copy text involved.

With an intermediate level of experience, I've had the opportunity to dive into both the front-end and back-end aspects of web development, including a hands-on understanding of DevOps.

My front-end experience is well-rounded with considerable knowledge in both Angular and React. I've built dynamic and responsive user interfaces using HTML integrated with Angular, appreciating the framework's powerful capabilities for single-page applications. In addition to Angular, I've utilized React for its flexibility and efficiency, building reusable components that provide a seamless user experience.

On the back-end side, my skills extend to the use of Django, a high-level Python Web framework that encourages rapid development and clean, pragmatic design. My familiarity with Python aids in crafting efficient, secure, and scalable server-side logic, further bolstering the performance of the applications I develop.

To bridge the gap between development and operations, I've ventured into the realm of DevOps. My experience includes managing AWS resources like EC2 for computing capabilities, Lambda for serverless architectures, and S3 Buckets for scalable storage. These tools have allowed me to maintain an efficient pipeline, automate repetitive tasks, reduce system discrepancies, and ensure the scalable and reliable deployment of web applications.

Through this balanced skill set, I'm able to take a holistic approach to web development, managing the lifecycle of an application from initial design and development through to deployment and maintenance. This intermediate experience in diverse aspects of web development equips me to tackle complex problems, and build intuitive, efficient, and scalable solutions.