Implementing an automatic grading system for handwritten answer sheets using a multi-agent framework streamlines evaluation, reduces manual effort, and enhances consistency. A multi-agent system <\/a> (MAS) consists of autonomous agents that extract information, grade answers, and suggest improvements. By leveraging Handwritten Answer Evaluation using Griptape, educators can automate grading, ensuring accuracy and efficiency. This approach allows teachers to focus more on personalized feedback and student development while maintaining fairness and reliability in assessments with Handwritten Answer Evaluation using Griptape.<\/p>\n This article was published as a part of the\u00a0<\/strong><\/em>Data Science Blogathon.<\/em><\/strong><\/a><\/p>\n Multi-Agent Systems (MAS)\u00a0are complex systems composed of multiple interacting intelligent agents, each with its own specialized capabilities and goals. These agents can be software programs, robots, drones, sensors, humans, or a combination thereof.\u00a0MAS leverage collective intelligence, cooperation, and coordination among agents to solve problems that are too complex for a single agent to handle alone.<\/p>\n MAS can adapt to changing environments by adding or removing agents, making them highly scalable for complex problem-solving. Decentralized control ensures continued system operation despite component failures. MAS can tackle large-scale tasks by combining the expertise of multiple agents, outperforming single-agent systems.<\/p>\n The core components of multi-agent systems include agents, which are autonomous entities with specific roles and goals, acting as the cognitive core of the system. Tasks represent specific jobs assigned to these agents, ensuring that their efforts are directed towards achieving the system\u2019s objectives. Tools extend the capabilities of agents, allowing them to interact with external systems and perform specialized tasks efficiently. Additionally, processes outline how agents interact and coordinate actions, ensuring that tasks are executed in harmony. The environment provides the context in which agents operate, influencing their decisions and actions. <\/p>\n Finally, communication protocols enable agents to share information and negotiate, fostering collaboration or competition depending on the system\u2019s design. These components work together to enable complex problem-solving and adaptability in multi-agent systems.<\/p>\n Multi-agent AI systems can be incredibly useful in a variety of applications across different industries. Here are some examples:<\/p>\n Griptape is a modular Python framework designed to build and operate multi-agent systems, which are crucial components of agentic AI systems. These systems enable large language models to autonomously handle complex tasks by integrating multiple AI agents that work together seamlessly. Griptape simplifies the creation of such systems by providing structures like agents, pipelines, and workflows, allowing developers to build business logic using Python and ensuring better security, performance, and cost efficiency.<\/p>\n With the increasing prevalence of online classes and various modes of education, there is a growing shortage of staff to evaluate students\u2019 exams. The slow pace of evaluation remains a major bottleneck in improving instructors\u2019 productivity. Teachers often spend a significant amount of time grading hundreds of answer sheets, time that could be better utilized for tasks like projects, research, or directly assisting students. This issue is particularly relevant as multiple-choice exams are not always effective in assessing a student\u2019s understanding of a subject. In this article, we will develop a multi-agent system designed to automatically grade handwritten papers.<\/p>\n Implementing a multi-agent system for automatic grading of handwritten answer sheets can significantly streamline the evaluation process for educators. This system utilizes specialized agents to extract relevant information from the sheets, assess the answers based on predefined criteria, and even provide suggestions for improved responses. By automating these tasks, teachers can focus on more critical aspects of education, such as personalized feedback and student development. This technology can also enhance grading consistency and reduce the time spent on manual evaluation.<\/p>\n We will build this system using GripTape on Google Colab with T4 GPU (Free Tier).<\/p>\n Automating handwritten answer evaluation with a multi-agent system can improve accuracy, efficiency, and consistency. By leveraging Griptape, educators can streamline grading, reduce manual effort, and ensure fair assessments.<\/p>\n The code below installs necessary dependencies for working with Griptape, Ollama, and Langchain, followed by importing various modules to facilitate creating and managing agents, tasks, and tools for handling different data types and web searches. It prepares the environment to execute a multi-agent system using AI models and external tools like file management and web search.<\/p>\n The following code starts the ollama server. We also pull \u201cminicpm-v\u201d model from ollama so that this vision model can be used to extract text from handwritten notes.<\/p>\n The code below defines the \u201cimage_dir\u201d that stores our images or handwritten answer sheets. Also a a function reading_answersheet is defined that initializes an agent with tools for managing files and querying images using a vision language model (\u201cminicpm-v\u201d). The agent uses a file manager and an image query tool to process images.<\/p>\n \u00a0 This code defines a function evaluation_answer that creates an agent with tools for web search using DuckDuckGo.\u00a0\u00a0<\/p>\nLearning Objectives<\/h3>\n
\n
What are Multi Agentic Systems?<\/h2>\n
Key Characteristics of Multi-Agent Systems<\/h3>\n
\n
What Constitutes Multi Agentic Systems?<\/h2>\n
Some Key Application Areas of Multi Agentic System<\/h2>\n
\n
GripTape For Building Multi-agent Systems<\/h2>\n
Core Components of GripTape<\/h3>\n
![\"Core](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/Core-components-of-GripTape.webp\")
<\/figure>\n\n
Our Solution: Agents and Tasks<\/h3>\n
![\"Agents,](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/Agents-Tools-Tasks-used-in-Our-Multi-Agentic-System.webp\")
<\/figure>\nHands on Implementation of Automatic Grading of Answer Sheets<\/h2>\n
Step 1: Installing and Importing Necessary Libraries<\/h3>\n
!pip install griptape\n!sudo apt update\n!sudo apt install -y pciutils\n!pip install langchain-ollama\n!curl -fsSL https:\/\/ollama.com\/install.sh | sh\n!pip install ollama==0.4.2\n!pip install \"duckduckgo-search>=7.0.1\"\n\n\nimport os\nfrom griptape.drivers.prompt.ollama import OllamaPromptDriver\nimport requests\nfrom griptape.drivers.file_manager.local import LocalFileManagerDriver\nfrom griptape.drivers.prompt.openai import OpenAiChatPromptDriver\nfrom griptape.loaders import ImageLoader\nfrom griptape.structures import Agent\nfrom griptape.tools import FileManagerTool, ImageQueryTool\nfrom griptape.tasks import PromptTask, StructureRunTask\nfrom griptape.drivers.structure_run.local import LocalStructureRunDriver\nfrom griptape.structures import Agent, Workflow\n\nfrom griptape.drivers.web_search.duck_duck_go import DuckDuckGoWebSearchDriver\nfrom griptape.structures import Agent\nfrom griptape.tools import PromptSummaryTool, WebSearchTool<\/code><\/pre>\nStep 2: Starting Ollama Server & Pulling the Model<\/h3>\n
import threading\nimport subprocess\nimport time\n\ndef run_ollama_serve():\n subprocess.Popen([\"ollama\", \"serve\"])\n\nthread = threading.Thread(target=run_ollama_serve)\nthread.start()\ntime.sleep(5)\n\n!ollama pull minicpm-v\n<\/code><\/pre>\nStep 3: Setting OpenAI API Key<\/h3>\n
import os\nos.environ[\"OPENAI_API_KEY\"] = \"\"<\/code><\/pre>\nStep 4: Creating an Agent to Read Handwritten Answer Sheets<\/h3>\n
images_dir = os.getcwd()\n\n\ndef reading_answersheet():\n driver = LocalFileManagerDriver(workdir=images_dir)\n return Agent(\n tools=[\n FileManagerTool(file_manager_driver=driver),\n ImageQueryTool(\n prompt_driver=OllamaPromptDriver(model=\"minicpm-v\"), image_loader=ImageLoader(file_manager_driver=driver)\n ),\n ]<\/code><\/pre>\nStep 5: Creating an Agent For Evaluation of Answers using Web\u00a0<\/h3>\n
def evaluation_answer():\n return Agent(\n tools=[WebSearchTool(web_search_driver=DuckDuckGoWebSearchDriver()), PromptSummaryTool(off_prompt=False)],\n )<\/code><\/pre>\nStep 6: Creating Tasks For Reading Handwritten Answers, Scoring them and Suggesting Improvements<\/h3>\n
![\"sample1:](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/sample1.webp\")
<\/figure>\n![\"sample1](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/image_93.webp\")
<\/figure>\n![\"sample1](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/chemistry_1.webp\")
<\/figure>\n![\"sample](\"https:\/\/cdn.analyticsvidhya.com\/wp-content\/uploads\/2025\/03\/image_94.webp\")
<\/figure>\n![\"Nibedita](\"https:\/\/av-eks-lekhak.s3.amazonaws.com\/media\/lekhak-profile-images\/converted_image_ZkJo4gb.webp\")
<\/p>\n