Ivan Baravoi

- Simplified stack (self-hosted solutions instead of cloud ones), restructured the platform (moved from an expensive SPA hosting to cheaper VPS, while migrating to microservices) and optimized costs by removing unused and excessive services (from 100% to about 2% of the usual monthly upkeep and 98% in monthly savings) - Took ownership of technical direction and engaged in addition of more than 10 new features, including 3 external integrations; dynamic content rendering based on web storage contents; email service and RBAC; DevOps and CI/CD processes; interactive lectures, homeworks and exams - Contributed to product planning, content creation and design decisions by clarifying, formalizing and structuring offers, payments, content, and challenging designs. Took part in AI tools introduction for business processes and development by orchestrating a framework, which led to almost 10 times growth in development speed. - Accelerated development and execution, driving the product to beta stage in 3 months

I have been tasked with finishing the web platform for a startup in EdTech. The current stack is JS/TS, Next, React, Node, Supabase. Primary focus of the platform is to assist Swiss children with studies and with preparation to a set of exams. Platform gives access to a lot of content ranging from webinars and course videos to interactive tasks for children, as well as task constructors for teachers.

- Manage software development projects end-to-end, including cost and time estimations, task planning, and progress tracking. Oversee multidisciplinary teams, set measurable goals, and supervise their successful completion within deadlines and budget. - Assess the technical skills of candidates and guide their integration into development teams. - Write technical tasks for developers; technical documentation and specification for developed software; manuals for the end users. - Mentor students and help to defend graduate and undergraduate theses; conduct code reviews and onboarding sessions for junior developers; help candidates acquire job-related skills in IT and receive job offers. - Teach Algorithms, Data Structures, Full-Stack and Desktop Software Development to university students.

Data Science & Machine Learning - Developed Agentic and RAG systems based on ChatGPT, IBM Watson, LLAMA, DeepSeek, and Claude for a variety of tasks, including data analysis, programming, and technical content generation. - Built an autonomous driving trajectory prediction model with automated GPU provisioning and cloud-based training workflows. - Developed NLP-focused ML/DL algorithms from scratch, including CNNs and RNNs. - Embedded web scraping algorithms in data pipelines and web applications with libraries requests, aiohttp and beautiful soup. - Work daily with Python data tools, such as numpy, pandas, matplotlib, plotly, seaborn, scikit-learn, keras/tensorflow, pytorch, statsmodels, prophet, scipy, airflow, mlflow, cometml, wandb, lightgbm, xgboost, catboost, nltk, spacy, gensim, transformers, Hugging Face. Web Development - Built full-stack e2e web applications using React (frontend), Django/FastAPI or Node.js/Next.js (backend), PostgreSQL/SQLite (databases), and AWS/GCP/Azure/Yandex (cloud). - Converted Figma designs into responsive, production-ready interfaces. - Integrated and created APIs for third-party services (Telegram, OpenAI, Google Docs, etc.). - Maintained and optimized legacy WordPress (PHP) codebase to support modern analytics and client acquisition tools. Desktop & Systems Programming - Built GUI applications with Tkinter (Python) and Qt (C++/Python). - Designed advanced CLI tools in C++ using OOP, metaprogramming, templates and design patterns.

◦ Lead the server-side redevelopment of a complex web platform (youngjunior.ru), transforming a fragmented legacy codebase into a scalable, maintainable backend and operational infrastructure that supported early adoption and scaling with zero obstacles (30k+ DAU and 5k+ registrations on the platform within the first 90 days; 10k in 180 days) ◦ Designed and implemented the new system architecture from the ground up, introducing clear modularization, robust coding standards, and comprehensive documentation ◦ Established CI/CD pipelines and defined best practices for code quality, deployment workflows, and team collaboration ◦ Oversaw technical operations and networking configuration through a custom bug-tracking/tracing system, ensuring reliable service delivery and smooth integration with third-party systems, as well as limiting downtime to a single quickly resolved issue ◦ Managed cross-functional coordination with other contractors and stakeholders to align priorities, requirements, and timelines ◦ Mentored junior developers through pairing sessions, code reviews, and project-based initiatives, significantly raising team competency and retention ◦ Acted as a supporting organizer for a frontend coding contest: contributed to technical task development, provided participant consultations during the event, helped resolve bugs and facilitated implementation of new features in real time ◦ Finished adopting the agile development, formed a steady stream of tasks to backlog in close collaboration with PM, set up urgent infra support and fallback system

◦ Searched for valuable market information using open-source data ◦ Applied financial, fundamental, and technical analyses to specify trade ideas from gathered information ◦ Developed predictive models and built trading strategies using Python, R, C++, and mathematical methods, such as regression, autoregression models (SARIMAX, Prophet, etc), and statistical analysis ◦ Presented trading strategies to the Senior Analyst

▪ 1400+ hours in Math including probability theory and statistics, numerical, calculation and optimization methods, PDEs etc ▪ 180 hours in Computer Science including introduction to C ▪ 280+ hours in Economics including mathematical methods applications

◦ Learned about libraries and ML/DL algorithms used in Chemistry for Rational Design and Automated Discovery purposes ◦ As a graduation project resembled research from a scientific paper in chemistry, where molecular structures were translated to numerical fingerprints and used for training random forest models to predict reaction results

Got introduced to C/C++, SQL, Python and some ML libraries, such as numpy, pandas, scipy, matplotlib, scikit-learn etc.