Denys T.

Platform that develops very functional human-like LLM agents that can interfere with the real world and execute tasks.

Responsibilities:

I was working on a platform that develops functional human-like LLM agents that can interfere with the real world and execute tasks.Was designing and implementing a pluggable, async (event-driven) architecture for integrations (WhatsApp, email, streamlet forms, Web UI, streaming intermediate responses from agents, etc.) that works with LangChain agents.

As a full-stack developer was part of the team that creates the application that helps medical staff to communicate, organize their work, share documents, create events, facilitate the assessment of interns, etc. (very popular in Australia and New Zealand).

Responsibilities:

• Developing REST API on top of Aiohttp framework, working with MongoDB database and Celery + RabbitMQ asynchronous tasks. Also have created web client on top of ReactJS and Redux for this application.
• Developing the application for team-based work with tasks (something like Asana but simpler). Server side of the application was on top of Flask and FlaskRESTfull. As a database, It has used PostgreSQL and SQLAlchemy ORM.
• Creating predictive models for time series data with recurrent neural networks (on Keras and Tensorflow). Also data preparation with pandas and scikit-learn and some data engineering.
• I was creating complete pieces of functionality for end users (starting from requirements, doing both design and implementation), participating in system design decision-making as well as supporting and improving existing features, performance optimization, code reviews, and refactoring
• I was working with frameworks and libraries such as ReactJS, and Redux, and had a little bit of experience with TypeScript on the frontend side as well as Python/Flask, Postgres, and AWS lambdas on the backend side.
  ‍

During this time developed some crucial features of the applications from start to finish, like the creation and management of events for doctors and interns, a scheduled notifications system for many other features, a significant part of interns' assessment process system, etc, as well as event tracking, Cypress tests integration and others.