About the project
Many college students can read and write English but freeze when they have to speak it in an interview, a group discussion or a simple phone call. Spoken English classes are costly and give each learner very little speaking time. AI Spoken English Tutor (BolChaal) solves this with an Android app that acts as a patient speaking partner: the student talks, the AI replies by voice, and every answer is checked for grammar, vocabulary, fluency and pronunciation. A trainer follows each student's progress, and an admin controls topics, users and AI settings from a web panel.
How it works
- The student registers and takes a 5-question AI placement test that sets the level (Beginner, Intermediate or Advanced).
- The student picks a topic such as Job Interview, Ordering Food or At the Railway Station; the AI plays the role (HR manager, waiter, ticket clerk).
- Android speech-to-text turns the student's voice into text and the AI reply is spoken back with text-to-speech.
- At the end, Gemini returns scores, corrections and new words. If the internet or API key is missing, an offline rule-based checker gives the result instead.
- Pronunciation practice compares the spoken sentence word by word and marks wrong words in red.
- The weak-area report groups past mistakes (tenses, prepositions, articles) and gives a mini lesson with a 5-question quiz.
- XP, daily streak, badges, a batch leaderboard and a level certificate (PDF) keep students practising.
Users
- Student: placement test, AI conversation practice, scores and corrections, pronunciation, vocabulary, weak areas, challenges, progress chart, certificate.
- Trainer: sees their students, reviews sessions with star rating and feedback, adds topics, assigns topics with a due date and views a batch report.
- Admin: approves trainers, manages students, categories, topics, word bank, badges, notices, AI settings and printable reports.
Highlights
- 3 roles, 26 Android screens, 33 PHP APIs, 16 admin pages, 14 MySQL tables
- Gemini API for conversation, scoring, placement test and lessons, with a full offline fallback
- Speech-to-text and text-to-speech with accent and voice-speed settings
- Line charts, progress timeline, leaderboard and PDF certificate / session report
- Validations on every form in the app and on the server (email, 10-digit mobile, password, duplicates, ranges, dates)
- Hashed passwords and prepared SQL statements
- Sample data with ready demo logins for student, trainer and admin
What you get
- Complete Android Studio project (Java) and PHP admin panel + APIs
- MySQL database file with realistic sample data
- Project report (Word, 51 pages) with DFD, ER, UML diagrams and test cases
- Project abstract (Word + PDF), project explanation with viva questions
- Step-by-step setup guide (PDF) and README
- Setup help if you get stuck
Who can use this project
Suitable as a final year project for BE / B.Tech (Computer, IT, AI & DS), MCA, BCA and Diploma students. It is a good choice if you want an Android project with source code that uses AI, speech recognition and a real admin panel, and that comes with documentation and a report ready for submission.
Frequently asked questions
Does the app work without a Gemini API key? Yes. An offline checker scores sessions and runs the conversation with built-in replies, so the demo never stops. With a free Gemini key the feedback becomes much richer.
Will speech-to-text work on the emulator? Yes, if the emulator has the Google app and microphone access enabled. A real Android phone gives the best result.
Can I add my own topics? Yes. Trainers add topics from the app and the admin can add or edit topics, levels and goal words from the web panel.
Can I run it on a real phone? Yes. Keep the phone and PC on the same Wi-Fi and type the PC's IPv4 address on the login screen.
