Inspiration

The genesis of Evater stemmed from a common modern dilemma. I observed my aunt's concern over her children's engagement with mobile phones, even as one of them, despite minimal daily study (10-15 minutes), excelled in school. This performance, coupled with the child's logical query - "Why study more if I'm performing well?" - highlighted a critical flaw in our existing educational assessment methods.

It became clear that while learning methodologies have accelerated exponentially, evaluation techniques have remained largely stagnant for decades. This disparity begged for a solution. My objective was to address the unmet needs of all key stakeholders: students, parents, and teachers.

Through research, I identified a prevalent issue in Indian education: rote learning. This is often a direct consequence of teachers evaluating students on memorized answers rather than conceptual understanding, effectively turning students into "if/else machines." While numerous tools exist for self-directed learning, there's a significant gap in effective, real-time assessment and feedback mechanisms.

Active learning, paired with instant feedback, emerged as the crucial missing link. However, this is currently hampered by several factors:

  • Parents often lack the pedagogical expertise to provide effective, personalized feedback.
  • School children struggle to self-evaluate accurately and objectively.
  • Teachers face bandwidth constraints that prevent them from personalizing education for each student.

After extensive research and valuable conversations with end-users (students), the concept of Evater was born.

What Evater Does

Evater is designed to empower personalized learning by enabling the creation of custom tests for middle school students (Grades 6-10) across Science, Mathematics, and Social Science. Beyond test generation, its core functionality lies in evaluating responses and providing detailed, constructive feedback, mimicking the personalized guidance a dedicated teacher would offer.

How Evater Was Built

The project leveraged a modern technology stack for rapid development and scalable deployment:

Frontend: Developed using Flutterflow, chosen for its efficiency in building cross-platform applications.

Backend: Implemented in Python using FastAPI, providing a robust and high-performance API. This backend is served on Modal for serverless inference, ensuring scalability and ease of deployment.

Core Logic: Dspy was utilized for the underlying intelligence that powers the test generation and evaluation functions. Evater is powered, by gpt4.1 (for test generation and evaluation) and gemini 2.5 flash (for OCR)

Challenges Faced

Developing Evater presented a unique set of challenges, primarily due to this being my first foray into building a full-scale production application. Key areas that required significant learning and problem-solving included:

  • Setting up robust authentication mechanisms.
  • Understanding and implementing effective backend serving strategies.
  • Identifying and integrating free services to facilitate the initial Minimum Viable Product (MVP) rollout. While no single challenge was "stark," the cumulative effort of navigating these various small obstacles contributed significantly to the learning curve.

Accomplishments We're Proud Of

The most rewarding accomplishment has been the genuine adoption and positive reception from the target audience. My middle school cousins are actively using the app, and a particularly heartwarming moment came from my younger sister who, upon using Evater, remarked, "I won't share it with any of my friends, because they can score better than me." While perhaps not ideal for immediate viral growth, this comment was a profound testament to the perceived value and effectiveness of Evater from a student's perspective. It underscored that the product genuinely addresses their need for effective, personalized evaluation.

What We Learned

The journey from prototype to production underscored the significant complexities involved. Even minor oversights can lead to considerable friction in a live environment, emphasizing the importance of meticulous planning and execution across all development phases.

What's Next for Evater

The future of Evater is focused on expanding its reach and enhancing its capabilities:

  • Promotion and Accessibility: A key next step is to actively promote Evater, especially aiming to make it accessible to underprivileged children. We believe its intuitive UI and ability to provide tailored feedback, unlike generic AI tools requiring advanced prompting, could be a game-changer for them.

  • Grade Expansion: Extending support to all grades (1-12).

  • Subject Diversity: Adding more subjects, including languages, secondary subjects like IT and General Knowledge.

  • Diagram Support: A significant technical challenge involves adding support for generating and evaluating questions that incorporate diagrams.

  • Advanced Math Evaluation: Continuously improving the accuracy and depth of evaluating mathematical answer sheets.

Built With

Share this project:

Updates