Description

Project Title
AI French Tutor: An Adaptive Conversational Language Learning Platform

General Description
The global market for online language learning is dominated by gamified applications such as Duolingo and Babbel, which excel at vocabulary drills but fail to replicate the most valuable element of language acquisition: extended, personalized interaction with a fluent speaker. Human tutoring at platforms like Preply or Italki costs 15 to 40 USD per hour and does not scale. As a result, hundreds of millions of motivated learners are stuck between two unsatisfactory options.
This project aims to design and build an AI-powered French tutor capable of replacing the core experience of working with a private teacher: open-ended conversation, real-time correction of grammar and pronunciation, adaptive lesson planning based on individual weaknesses, and structured progress over weeks and months. The product is intended for a global English-speaking audience of adult learners studying French for work, study, immigration, or personal goals.
Students working on this project will tackle one of the most active research areas in applied AI today — building an autonomous tutoring agent that approaches the quality of a human teacher in a specific, well-defined domain. The goal is not to integrate existing chatbot APIs into a wrapper, but to push beyond what current AI tutors on the market are able to do.

Project Scope
The purpose of the project is to research, design, and implement a working AI tutor for the French language that performs better than the current generation of AI language tutors available to consumers. Students will define their own technical approach — the choice of models, frameworks, speech processing tools, and evaluation methods is intentionally left to the team.
The system, at minimum, must be able to:
•	Hold an extended, natural conversation in French with a learner at any level from A1 to C1, adapting register and complexity to the learner.
•	Listen to the learner speak and evaluate pronunciation, including features specific to French such as liaison, nasalization, and elision.
•	Identify recurring grammatical and lexical errors across sessions and adjust subsequent lessons to address them.
•	Build and maintain a personalized learning path for each user, predicting which topics or structures the learner is likely to struggle with next.
•	Provide actionable, specific feedback after each interaction — not generic praise, but the kind of correction a skilled teacher would give.
Students are expected to study existing solutions in the market, identify their weaknesses, and propose original improvements. The deliverable is a working prototype that demonstrates measurable advantages over the current state of the art on specific, defined criteria chosen by the team.
The project combines applied research with product engineering. Students will gain exposure to natural language processing, speech recognition and evaluation, adaptive learning algorithms, full-stack application development, and the methodology of evaluating AI systems against human benchmarks.
Student Requirements
•	Teamwork and collaboration in small groups.
•	Full attendance and active participation in weekly meetings.
•	High motivation and a problem-solving mindset — this is an open research and engineering problem, not a tutorial.
•	Independent learning and willingness to read recent research in NLP, speech processing, and adaptive learning.
•	Personal responsibility for assigned tasks and deadlines.
•	Comfort with ambiguity: the team chooses the technical stack, the evaluation methodology, and the architecture.
•	Knowledge of French is not required. The product is built for English-speaking learners- native French speakers on the team are welcome but not necessary.
Development Tools
The choice of technologies, programming languages, libraries, models, and APIs is left to the student team. The project is deliberately open in this regard. Students are expected to evaluate available options, justify their choices, and document their decisions.
Suggested categories the team will need to address (specific tools to be selected by the team):
•	A frontend framework for delivering the application across devices.
•	Speech recognition for transcribing the learner's spoken French.
•	Speech synthesis for generating the AI tutor's voice.
•	A large language model or combination of models for understanding, evaluating, and responding to learner input.
•	A method for evaluating French pronunciation quality.
•	A backend and database for storing user progress, error history, and adaptive learning state.
•	Version control (Git / GitHub) and a project management approach of the team's choosing.
Deliverables
The project team is expected to produce the following:
•	Specification document — including project plan, methodology, technical architecture, choice of tools with justification, and evaluation criteria for measuring success against existing AI tutors.
•	GitHub repository with the full codebase, including documentation sufficient for a third party to run and extend the system.
•	A working prototype of the AI French Tutor, accessible through a web interface, demonstrating the core capabilities described in Project Scope.
•	Comparative evaluation report — benchmarking the prototype against at least two existing AI language tutoring products on criteria defined by the team (e.g., conversational quality, pronunciation feedback accuracy, adaptiveness, learner outcomes).
•	Final presentation — slides summarizing the research approach, key technical decisions, results, and findings.
Additional Notes
•	Each project team may include up to 5 students.
•	French.Super provides domain expertise (DELF/DALF examiner methodology, ten years of teaching experience, eight published preparation guides), an audience of 90,000+ learners for beta testing, and full operational support.
•	The product has a clear path to real-world use after the project: the team's work will be evaluated by, and potentially deployed to, an existing audience of paying language learners.
•	Industry contacts for student inquiries:
Ilia Gavrilov — Co-founder, French.Super (project lead, methodology, content)

Email: elie.gavrilov@gmail.com
Sergei Safronov — Co-founder & CTO, French.Super (technical contact, AI infrastructure, product)
Email: safronovsergeisafronov@gmail.com
For any questions regarding the proposal, you may also contact: Shirany.hit@gmail.com
AI French Tutor: An Adaptive Conversational Language Learning Platform

Full Stack

Description

Emphasis in project execution

The project is has cooperation with the industry and combines meeting deadlines while being creative and focused on the task
