Week #5

Week #5 #

Feedback #

Sessions #

  • It will be greate to customize personal lessons plan from settings
  • The design of the app is pretty simple and looks good. But i think that it will be good to add picture to the background corresponding to the learning goal
  • The word database is huge and really helpful, but more variety in exercises would make lessons even better
  • It will be good to add pronounce training to words from dictionary and custom lessons for words from non-learned group
  • Loggin system is simple, but i think that it will be good to add other services for loggin

Analyze #

  • The customization of the lessons will be added to profile as soon as possible
  • Since some models of the lessons are not made yet, we are going to implement them at the next week
  • The idea of the repeating words from dictionary is good one, we will discuss about it

Iteration & Refinement #

Implemented features based on feedback #

  • Discuss profile settings that we will have
  • Discuss exercise versions
  • Discuss word repeatitions

Performance & Stability #

==================================================
📊 IMPORT STATISTICS
==================================================
⏱️ Duration: 1h24m43s
🚀 Speed: 6.8 rows/second
📄 Total rows processed: 34743
📝 Words imported: 34743
🏷 Topics created: 570
💬 Sentences added: 172060
❌ Errors encountered: 106

Documentation #

Project README

iOS README
Thesaurus README
CONTRIBUTION.md

All docs

ML Model Refinement #

Thesaurus

  • 3.5 MB RAM
  • 7.5 ms per request
  • VRAM: 0.55 GB

Weekly commitments #

Individual contribution of each participant #

Danila Kochegarov (Team Lead & Backend Developer) #

  • Fixed filling script
  • Continue developing telegram bot functionality
  • Fixed issues with lesson
  • Connect ML with backend
  • Integrate ML into database filling
  • Wrote script that access website with word transcription for filling database

Savva Ponomarev (iOS Developer & Project Manager) #

  • Wrote report
  • Add navigation bar and new screens:
    • Calendar with statistic of the lessons
    • Statistic screen with information about progress
  • Integrate new endpoints from backend to app
  • Started to work with cahcing info

George Selivanov (System Analyst) #

  • I found and tested many different llm
  • Tested for specific requests for working with English in dialog format
  • Wrote a fault-tolerant microservice for interaction with the service for selected models of the groq and gemini service

Timofey Ivlev (DevOps Engineer) #

Anton Korotkov (ML Engineer) #

  • Improving the distractors creation: now supports different forms of words
  • Fixed random errors occuring during working of NLP module.

Daniil Tskhe (Backend Developer) #

  • Wrote the logic for saving the user’s progress after the lesson
  • Fixed the paths for the router
  • Fixed auth logic and models creation error (user pref before user = error, same with sentence before word = error). Very important for database, where we need to see right relationship

Evgeniy Bortsov (Android Developer) #

  • Statistic screen
  • Statistic calculation locally
  • Fix routing between Home screen and lesson screen
  • After lesson the result sending back to the backend

Plan for Next Week #

  • Complete any remaining high-priority features or bug fixes
  • Ensure all code is well-commented and clean
  • Documentation Finalization: Ensure README.md is comprehensive (project overview, features, tech stack, setup instructions, deployment link)
  • Presentation Preparation: Develop slides covering the problem statement/solution, target audience, key features/tech stack, live demo, challenges, future work, team contributions
  • Finalize API documentation

Confirmation of the code’s operability #

We confirm that the code in the main branch:

  • In working condition.
  • Run via docker-compose (or another alternative described in the README.md).