Daniel – Python, Django, Golang

The main goal of the project was to develop a Trade & Supply Hub for oil market traders to support their decision-making process. The initiative aimed to protect and grow approximately $500 million in annual value by leveraging the full Trade & Supply information flow in Crude and Products.

Additionally, I served as the lead developer on a related project that provided Trade & Supply offices with an API, enabling them to access T&S intelligence through Excel, Python, or JavaScript.

• Designed and implemented a RESTful API providing unified product data across all applications within the Geneva initiative.
• Analyzed project requirements and proposed the adoption of a modern framework (FastAPI) instead of the one initially suggested by the department lead.
• Contributed to defining a data-driven architecture with the API as the core component of the system.
• Implemented a lightweight LLM-based solution to enable dataset discovery within the Trading & Supply domain.
• Assisted in establishing the team’s management framework in collaboration with an entry-level project manager.
• Acted as the team’s DevOps engineer, managing CI/CD pipelines and infrastructure processes.
• Applied Terraform and Azure expertise to deliver end-to-end solutions, including development, testing, and deployment.
• Ensured product security and quality by maintaining documentation, implementing static and dynamic code analysis, and developing both unit and integration test frameworks.
• Introduced a new authentication approach using OpenIDC (PingID federation) and developed a standalone authorization library shared across the client’s environment.
• Mentored an entry-level backend developer; actively contributed to code reviews, branching strategy decisions, and toolset selection.

Global solution for managing climate risk and measuring temperature alignment across investment portfolios — down to the individual asset level. The project consisted of two key modules: a domain-specific language (DSL) for model definition and execution, and a set of climate change scenario models designed to assess financial exposure and temperature alignment metrics.

• Contributed to the development of a highly declarative Domain-Specific Language (DSL) operating at the Abstract Syntax Tree (AST) level.
• Extended the language by introducing new data types (Date, DateTime), enhancing syntax, adding mathematical functions (logn, abs), and implementing data models defined by analysts directly within the language.
• Maintained and optimized the runtime execution layer to improve system performance.
• Collaborated with data analysts and stakeholders to enhance model implementation and usability.
• Improved data ingestion by adding support for loading equality and security bonds data from CSV and Bloomberg-formatted files.
• Supported team growth by onboarding new members and delivering technical presentations about system architecture and design.
• Served as the “fix lead,” providing first-line user support for data analysts and stakeholders, managing incidents and coordinating resolutions across multiple teams.

A platform for human-in-the-loop improvement of speech recognition algorithms, enabling users to manually correct transcribed text, annotate sentiment, and provide other linguistic feedback to enhance model accuracy.

• Developed a RESTful API in Python for a web-based speech recognition platform.
• Maintained WebSocket client and server components alongside a CRUD solution integrated with a Go-based HTTP communicator used by an external application.
• Took ownership of a legacy Go codebase and refactored the components responsible for exposing the HTTP API to a mobile speech recognition application.
• Created and standardized missing API documentation using Postman to improve maintainability and onboarding efficiency.

A social media application developed for the Californian fashion industry, inspired by Instagram. The platform replicated key Instagram features such as photo sharing, likes, and comments, while introducing a custom-built administration panel tailored to the client’s business needs.

• Implemented a backend API for a social media application inspired by Instagram.
• Developed user and article modules as core parts of the main API.
• Built a fully custom moderator panel using Django, Django REST Framework, and jQuery, enabling user management (blocking, highlighting “photo of the day,” posting global commercials, etc.).
• Designed and implemented an algorithm to group user activities by time and type.
• Documented all APIs using Swagger for consistent and accessible developer references.
• Supported the QA process by performing manual and Postman-based API testing.

An e-commerce and production management system developed for a metallurgical company. The platform was designed to streamline production scheduling, order placement, and employee management, integrating operational and business processes into a single solution.

• Developed the ordering module, including full CRUD operations and automated stock updates.
• Integrated a third-party invoicing service to generate and email invoices automatically via a custom queue-based messaging system (Celery + RabbitMQ).
• Implemented delivery service integration using a SOAP interface, synchronizing shipment data with the system’s SQL database.
• Assisted the front-end developer with the implementation of smaller features and UI adjustments.
• Collaborated closely with the client’s representative to resolve issues caused by the team’s limited experience.
• Proposed and executed a plan to modularize the application, addressing problems incrementally and improving stability.
• Participated in manual testing and troubleshooting to ensure product reliability and smooth operation.