Bijaya – Python, Typescript, PostgreSQL

Created a sophisticated hotel booking assistant harnessing the power of Large Language Models (LLM). Applied vector embeddings to enhance data representation and retrieval efficiency. Implemented the Retrieval Augmented Generation (RAG) framework alongside LLM architecture to optimize conversational interactions and offer personalized recommendations to users.

• Created LLM agent assistant;
• Implemented RAG architecture;
• Built a pipeline for vector embeddings;
• Utilized scalable vector database;
• Advanced vector search.

Helping national banks, airlines, telcos, pharmaceutical companies, government, and the public sector deploy custom and commercial apps remote to 1000s of staff around the globe.

• Led the successful development of the latest iteration of ReFRMD, demonstrating adept leadership in overseeing the project from inception to completion.
• Innovated an abstract syntax tree-like syntax tailored for describing ReFRMD-specific form UI elements. This breakthrough allowed for the seamless integration of complex validation logic, including dependencies on dynamically specified UI elements.
• Engineered a cutting-edge solution enabling the dynamic construction of forms coupled with robust validation mechanisms and type safety throughout the development process.

Your journaling companion.

• Led the development of the MVP for a cutting-edge conversational AI, taking charge of cloud, backend, and frontend stacks.
• Pioneered the integration of LLM with other AI models, resulting in an AI that excels in natural language processing across text, audio, and speech interactions.
• Engineered a sophisticated AI capable of storing and analyzing personality traits, memories, goals, and preferences, enriching the conversational experience for users.

No-code IDE for Quantum Computing. It builds a bridge between the end-users and quantum computing with visual, easy-to-use software development tools. It has been widely adopted in many quantum computing subjects, including prestigious institutions like Oxford University and Stanford University. The platform received global recognition and was nominated for The Webby Awards 2023 in the science category, further highlighting its impact and success.

• Created a modern web application for quantum computing hardware overriding and customizing almost every default browser behavior, such as scroll, context menu, etc.;
• Added new features on top of the existing simulation engine in Rust WASM, which allowed a great deal of flexibility to the users;
• Developed complex mathematical functions including Blackman waveform, kaiser waveform, and Piecewise Cubic Hermite Interpolating Polynomial from scratch;
• Integrated functions into the UI and the simulation with SVG and numeric arrays, enabling real-time parameter changes without performance issues;
• Detected and fixed major UI/UX issues and hard-to-find bugs and ported the web app to run smoothly on smaller screen devices which improved the number of users and the number of quantum experiments growth in the platform by 30%.

It is a comprehensive logistics solution. As the solo developer, he acted as the Tech lead, gathering requirements and building both the back end using Python and the front end using React.

• Created front-end and back-end architecture;
• Successfully developed the project;
• Acted as the Tech lead, overseeing the entire development process;
• Conducted requirement-gathering sessions with the product owner to align project objectives;
• Covered code with tests.

Provided by PASQAL, one of Europe's leading quantum computing companies, it is a platform to make neutral-atoms programming accessible to everyone with zero code. The zero-code platform also provides an auto-generated code output in Python continuously as the quantum device, pulses, and atoms are added, updated, and removed.

• Created a modern web application for quantum computing hardware overriding and customizing almost every default browser behavior such as scroll, context menu, etc.;
• Added new features on top of the existing simulation engine in Rust WASM, which allowed a great deal of flexibility to the users;
• Developed complex mathematical functions including Blackman waveform, kaiser waveform, and Piecewise Cubic Hermite Interpolating Polynomial from scratch ;
• Integrated all functions into the UI and the simulation with SVG and numeric arrays, enabling real-time parameter changes without performance issues.

It is a workforce management software suite that monitors employees' work device activity, such as keyboard and mouse movements, currently using applications, and takes screenshots. An admin dashboard was provided that showed all the activities and productivity scores for each employee.

• Implemented a scalable and resilient back-end service in Python using FastAPI and Postgresql for the workforce management software suite, using microservice design principles, Docker containerization, and Kubernetes orchestration;
• Developed a desktop client application using Python, Pyside2, and Qt5 for a workforce management software suite with a focus on functional programming. Successfully ensured cross-platform compatibility on Darwin, windows, and Linux and integrated log syncing with Graylog service;
• Utilized Time series analysis techniques with TimescaleDB to analyze and visualize large datasets with the high temporal resolution, enabling accurate forecasting and trend identification. Leveraged advanced SQL queries and Python libraries such as Panda and NumPy to extract insights and communicate findings to stakeholders.

It is a traveling and trekking planning app

• Developed a service-oriented monolithic backend, adhering to Domain-Driven Design principles, utilizing Python, AIOHTTP, and Postgresql to ensure efficient performance and scalability, which was deployed in AWS;
• Implemented Elasticsearch search functionality for efficient and accurate search results retrieval;
• Utilized PostGIS extension to efficiently perform GIS analysis and modeling efficiently, enabling accurate geospatial data management and processing.