Table of contents
AI Summary by Katonic AI
This blog details a 6-step process for building an AI-powered candidate scoring application using the Katonic AI Platform, requiring no coding skills. The solution automates CV screening by analysing applications against specific job requirements, assigning objective scores based on customisable criteria like technical skills, experience relevance, and project expertise.
Ever felt overwhelmed by stacks of CVs during your hiring process? You’re not alone. Recruiting teams across industries struggle to efficiently evaluate candidates against job requirements—especially when dealing with hundreds of applications.
But what if you could automate this process and get consistent, objective evaluations in seconds?
That’s exactly what we’ll explore today: how to build your own AI-powered candidate scoring application using Katonic AI Platform’s no-code capabilities. The best part? You don’t need to be a data scientist or developer to make it happen.
The Recruitment Challenge That Needs Solving
Before we dive into the technical bits, let’s talk about why this matters. Traditional CV screening is:
- Time-consuming (typically 23 seconds per CV, multiplied by hundreds of applications)
- Inconsistent (different reviewers apply different standards)
- Prone to unconscious bias (we’re all human, after all)
- Often misses qualified candidates buried in the pile
These challenges aren’t just frustrating—they’re costly. According to research, poor hiring decisions can
cost organisations up to 30% of the employee’s first-year earnings.
Enter AI-Powered Candidate Scoring
An automated scoring system analyses CVs against your specific job requirements, evaluating factors like:
- Technical skills match (25%)
- Experience relevance (25%)
- Project expertise (15%)
- Years of relevant experience (15%)
- Industry-specific experience (5%)
- And other customisable criteria
This gives each candidate an objective score, helping your recruitment team focus their attention on the most promising applicants first.
Building Your Application: A 6-Step Guide
Step 1: Create a Document Processing Agent:
First, you’ll need to create a Document Processing Agent in AI Studio on the Katonic Platform. This
specialised application type is designed to extract, analyse and summarise key information from
documents like CVs, contracts, and reports.
- Log into your Katonic AI Platform account
- Navigate to AI Studio
- Click “Create Project”
- Name your project (e.g., “resume-scoring”)
- Select project accessibility (Private or Shared)
- Choose “Document Processing” as your project type
This creates the foundation for your application, giving you access to powerful document processing
capabilities without writing a single line of code.
Step 2: Configure Resume Scoring Settings
Now you’ll need to configure how your application will evaluate candidates. This is where you set up the specific criteria and instructions for the AI:
- Provide clear instructions about how to evaluate candidates
- Define scoring categories (technical skills, experience relevance, etc.)
- Assign weights to different evaluation criteria
- Configure model parameters like temperature and token limit
The more specific your instructions, the more accurate your results will be. For example, you might
include guidance like:
As an experienced HR professional with 10+ years in technical recruitment, carefully
evaluate how well this candidate’s qualifications align with our job requirements. Provide a
concise assessment following these guidelines:
Evaluation Criteria (by percentage):
- Technical Skills Match (25%)
- Direct experience in required technologies/skills (20%)
- Depth of expertise shown through projects (15%)
- Experience Relevance (25%)
- Years of relevant experience (15%)
Step 3: Test and Deploy Your Application
Before going live, it’s crucial to test your application with a few sample resumes to ensure it’s scoring
accurately:
- Upload test CVs to verify processing accuracy
- Review the evaluation results
- Make adjustments to your configuration if needed
- Once satisfied, deploy the application
During testing, you’ll see a detailed breakdown of scores across your defined categories, along with
explanations for each rating. This transparency helps you refine the system and builds trust in the results.
Step 4: Retrieve API Details for Integration
After deployment, you’ll need to grab the API details to connect your application with other tools:
- Click on the API button in the deployed application
- Copy the API URL and access token
- Keep these secure – they’re your keys to accessing the service
These credentials will allow your Streamlit application to communicate securely with the document
processing service you’ve just created.
Step 5: Build Your Streamlit User Interface
Now comes the fun part – creating a user-friendly interface for your application:
- Create a GitHub repository for your project
- Develop a Streamlit application with these key components:
- Job description input area
- CV/Resume file uploader
- Entity extraction options
- Results display section
The code for this part is straightforward. For example, to create a text input for job descriptions:
python
# Render the job description section
st.markdown(job_description_section_style, unsafe_allow_html=True)
overall_query = st.text_area(" Enter the job description you want to evaluate the candidate
And for the CV upload functionality:
python
# Render the job description section
st.markdown(resume_section_style, unsafe_allow_html=True)
main_uploaded_files = st.file_uploader(" Upload a PDF of the candidates resume/CV.", accept_
You can also add features to extract specific information from CVs:
python
if st.button("Add Entities to Extract", ) or st.session_state['is_submitting']:
tb = st.text_input(" Insert entites to extract from the resume", key="save_entity", on_change
st.session_state['is_submitting'] = True
Step 6: Process and Display Results
Finally, you’ll process the uploaded files and display the results in a user-friendly format:
- Send the job description and CVs to your API
- Process the responses
- Display the results in a clear, actionable format
- Include options for sorting or filtering candidates
This gives recruiters an instant overview of how candidates measure up against the job requirements,
with detailed breakdowns of strengths and weaknesses.
The Business Impact
Implementing this AI-powered candidate scoring system delivers several key benefits:
- Time savings: Reduce screening time by up to 75%
- Consistency: Apply the same evaluation criteria across all applications
- Objectivity: Minimise unconscious bias in the initial screening process
- Improved quality of hire: Focus human review on truly qualified candidates
- Better candidate experience: Provide faster responses to applicants
One customer using this solution reported processing over 500 applications in under an hour – a task
that previously took their team nearly a week.
Video Walkthrough: Building Your Candidate Scoring Application
Below you’ll find a comprehensive video demonstration that walks through each step of creating your candidate scoring application on the Katonic AI Platform. This visual guide complements the written instructions above and shows exactly how to navigate the Katonic AI Studio interface.
Getting Started
Ready to transform your recruitment process with AI? The Katonic AI Platform makes it accessible even if you don’t have a technical background. Our no-code AI Studio lets you build, test and deploy sophisticated AI applications without writing complex code, while providing the enterprise-grade security features required for handling sensitive HR data.
