Built for a Friend: ServiceNow Interview Buddy Using Ollama and Llama 3.2

This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend

ServiceNow Interview Buddy

What I Built

I built ServiceNow Interview Buddy, an AI-powered interview preparation tool designed for ServiceNow professionals.

The idea came from helping a friend prepare for ServiceNow interviews. While reviewing potential interview topics together, I realized that finding role-specific questions that matched both the target role and experience level often required searching multiple resources and manually organizing relevant content.

As someone who regularly participates in ServiceNow technical evaluations and interview discussions, I saw an opportunity to create a lightweight AI-powered assistant that could instantly generate realistic interview questions tailored to a candidate's role and experience level.

The application allows users to select:

• ServiceNow Administrator
• ServiceNow Developer
• ServiceNow Consultant
• ServiceNow Architect

and specify their years of experience.

The tool then generates customized ServiceNow interview questions using a locally running AI model.

Demo

Application Screenshots

Landing Page

Role Selection

Generated Questions

GitHub Repository

View the GitHub Repository

Code

The complete source code is available in the GitHub repository.

Key project files include:

• app.py
• requirements.txt
• README.md

How I Built It

Technology Stack

• Python
• Streamlit
• Ollama
• Llama 3.2 (Open-weight AI Model)

The user interface was built using Streamlit.

Ollama was used to run the open-weight Llama 3.2 model directly on my laptop.

When a user selects a ServiceNow role and years of experience, the application dynamically generates a prompt and sends it to the local Llama 3.2 model, which then produces tailored interview questions.

The workflow is:

1. User selects a ServiceNow role.
2. User enters years of experience.
3. The application builds a dynamic prompt.
4. Llama 3.2 generates interview questions.
5. The questions are displayed through the Streamlit interface.

Everything runs locally without requiring any cloud-hosted AI service.

Why Does Open Innovation Matter?

Open innovation was a key factor in making this project possible.

Using an open-weight model provided several important benefits:

• No API costs
• No subscription fees
• Full control over the model
• Local execution
• Better privacy
• No vendor lock-in

Because the model runs through Ollama, the application can continue functioning even without internet access once the initial setup is complete.

A closed AI service would have introduced recurring costs and external dependencies. Open-source AI allowed me to build a practical and accessible solution that anyone can run on their own hardware.

Feedback From My Friend

The feedback was positive because the generated questions adapted to both the selected role and years of experience.

Instead of relying on static interview question lists available online, the application generated questions that felt more relevant to the selected ServiceNow role and experience level.

This helped make interview preparation more focused and highlighted areas that deserved additional attention before an interview.

Challenges

One of the challenges was learning how to connect a Streamlit application with a locally running AI model through Ollama.

Another challenge was ensuring that the generated questions remained relevant across different ServiceNow roles while keeping the application simple and easy to use.

My Agent Session

I used Microsoft 365 Copilot together with local development tools to plan, build, troubleshoot, document, and prepare this project for submission.

During development I:

• Installed and configured Ollama
• Downloaded and tested the Llama 3.2 model
• Built the Streamlit user interface
• Connected the application to a locally running AI model
• Created project documentation
• Prepared the GitHub repository and screenshots
• Drafted and refined the submission article

This demonstrates how AI-assisted development can accelerate the creation of practical applications while still requiring design, implementation, testing, and iteration by the developer.

Future Enhancements

Planned improvements include:

• Interview answer evaluation
• Candidate scoring
• ServiceNow certification preparation
• Learning recommendations
• Mock interview simulation
• Expected answer generation
• Personalized preparation plans

Conclusion

Building ServiceNow Interview Buddy was an excellent opportunity to combine open-source AI with a real-world professional challenge.

What started as an effort to help a friend prepare for ServiceNow interviews evolved into a practical tool that generates role-specific interview questions for administrators, developers, consultants, and architects based on their experience level.

Using Ollama and the open-weight Llama 3.2 model made it possible to build the solution without API costs while ensuring that everything runs locally and remains under the user's control.

This project demonstrates how open innovation can be used to create practical learning and career-development tools. It also provides a strong foundation for future enhancements such as interview answer evaluation, scoring, certification preparation, and mock interview simulations.