Interview: How I Built an AI Voice Job Screening Application
- Esuabom Dijemeni
- 3 days ago
- 6 min read
How a simple idea became an AI Voice Screening assistant that helps hiring managers screen candidates more efficiently
In an AI-driven world, hiring the right people remains one of the biggest challenges for hiring managers and recruitment agencies. Reviewing hundreds of CVs, scheduling countless phone screening calls, and ensuring every candidate receives a fair and consistent first interview is both time-consuming and resource-intensive.
What if the initial screening process could be automated without compromising the candidate experience?
To explore this idea, we spoke with Esua, an AI Product Builder, about how he designed and built an AI Voice Screening Assistant that automates the first stage of the recruitment process.
In This Interview
Why London Agentic AI Hack Night?
Interviewer: Why did you decide to attend the London Agentic AI Hack Night?
Esua:
I had four reasons for attending.
Get My Hands "Dirty" - hands-on experience building with Agentic AI and AI Voice technology
Explore New Technologies - experiment with platforms like Bimpe AI and Vonage.
Build Something Cool - challenged myself to take an idea from concept to a working prototype in just 60–90 minutes
Connect with Fellow AI Product Builders - meet and collaborate with other AI Product Builders, developers, and data professionals from the Data/BI Analyst London community
Problem Discovery
Interviewer: What problem were you trying to solve?
Esua:
Every hiring manager faces the same challenge: finding the right candidate in a sea of applications.
Today, recruiters often receive hundreds of CVs for a single role, many of which have been generated or heavily enhanced using AI.
Only a small percentage of applicants genuinely meet the role's requirements, making it difficult to identify the best candidates quickly.
The first phone screening plays a crucial role in the hiring process.
The challenge is that these conversations are repeated over and over again for every applicant.
Questions like:
Tell me about yourself.
Why are you interested in this role?
What experience do you have?
What salary are you looking for?
When can you start?
These conversations are essential, but they're also repetitive and time-consuming.
More on problem discovery - https://www.itydata.com/posts/product-loop-statement
Understanding the Users
Interviewer: Who is AI Voice Job Screening actually for?
Esua:
There are really two primary users.
The first user is the hiring manager / recruiter.
Their goals are to:
Spend less time conducting repetitive first-stage screening calls.
Identify the strongest candidates more quickly.
Reduce the overall time-to-hire.
Eliminate wasted effort reviewing and screening unsuitable applicants.
Focus more of their time on high-value interviews and hiring decisions.
The second user is the candidate
Their goals are to:
A fair and consistent screening process.
A simple, structured interview experience.
Immediate engagement instead of waiting days for an initial phone call.
Meaningful feedback after completing the screening interview.
Understanding both users are critical
More on understanding the Users - https://www.itydata.com/posts/product-discovery-framework
The Product Idea
Interviewer: So what did you decide to build?
Esua:
We [Team Fantastic CV] decided to build an AI Voice Job Screening Application that automates the first stage of the recruitment process.
Instead of recruiters conducting rushed phone screenings—or, in some cases, skipping them altogether—candidates have a natural conversation with an AI interviewer.
The application performs the entire first-stage screening by:
Reviewing the candidate's CV before the interview.
Calling the candidate automatically.
Asking structured interview questions tailored to the role.
Probing deeper when it identifies gaps in skills, experience, or responses.
Assessing and scoring the interview against predefined criteria.
Generating a concise interview summary with key insights and recommendations for the hiring manager.
Everything happens through an automated voice conversation.
The goal isn't to replace recruiters—it was to free them from repetitive screening calls so they could spend more time engaging with the most promising candidates.
More on product idea - https://www.itydata.com/posts/idea-framework
Build & Experiment - AI Voice Job Screening
Interviewer: How did you build the AI Voice Job Screening Application?
Esua:
We only had 90 minutes to build a working prototype, so we approached the challenge with an experimentation mindset rather than trying to build a perfect product.
Our hypothesis was simple: validate the idea as quickly as possible and learn whether an AI voice interviewer could successfully conduct a first-stage screening interview.
To accelerate development, we used Bimpe AI as the foundation of the application. It enabled us to rapidly experiment with AI voice technology, iterate on prompts, and test the candidate experience in real time.
The application was built using several core capabilities:
Workflows – Configured AI agents to manage the end-to-end job screening process.
Playground – Tested prompts, conversation flows, and interview scenarios in real time.
Conversations – Captured and reviewed the complete history of each AI voice interview.
Integrations – Connected the AI agent to external business systems and data sources.
Deployment – Published the AI interviewer across voice communication channels for testing.
By the end of the 90-minute build session, we had created a Minimum Viable Product (MVP) capable of conducting automated voice screening interviews, evaluating candidate responses, and producing structured interview summaries for hiring managers.
More on product experimentation - https://www.itydata.com/posts/product-experiment-framework
Retrospective & Lessons Learned
Interviewer: Taking a step back, what are your biggest takeaways?
Esua:
I like to reflect on every product journey rather than focusing only on what we built
Three lessons stood out from this experience.
Turn Up—Even When You're Tired or Nervous
It's easy to find reasons not to attend a hack night or community event.
I'm really glad I pushed myself to show up.
If I hadn't, I would never have built an AI Voice Job Screening application in just 90 minutes or met such an inspiring group of AI builders.
Explore New AI Possibilities
Before this event, I had very little experience building AI voice applications.
Using Bimpe AI completely changed my perspective.
In just 90 minutes, I went from having no real understanding of AI voice development to building a working AI Voice Screening prototype.
Connecting with Other AI Product Builders
Building AI products can be a surprisingly lonely journey.
Many of us spend hours experimenting, coding, testing prompts, and solving problems on our own.
Looking back, the biggest takeaway wasn't just the application we built. It was the mindset of experimenting, learning quickly, and building alongside others. Those are the experiences that accelerate both product development and personal growth.
More on product retrospective - https://www.itydata.com/posts/product-retrospective-framework
What's Next?
Interviewer: So, what's next?
Esua:
For me, the next step has two parts: making the next product decision and repeating the Product Loop.
Every product experiment ends with a decision. Based on my Five Types of Product Decisions framework, I've decided to persevere with building AI Voice solutions using Bimpe AI and Python.
My next Product Loop will focus on building a Python-based AI Voice Job Screening application that delivers an even more natural, human-like interview experience.
I'll be exploring richer conversations, smarter follow-up questions, improved candidate evaluation, and deeper integrations with recruitment platforms.
So, this isn't the end of the journey. It's the beginning of the next iteration.
More on product decision - https://www.itydata.com/posts/product-decision-framework
More on repeating the product loop - https://www.itydata.com/posts/repeat-the-product-loop
Stay tuned for my next interview, where I'll share what I built, what I learned, and how the product evolved.
Useful Links
🤝 Join our AI Product Builder Community – Connect with other builders, share ideas, and learn from real AI product journeys.
🛠️ Discover our AI Product Development Frameworks – Learn the proven frameworks we use to take AI products from idea to launch.
💡 Turn your idea into an AI app – Explore our AI App Development Service and transform your idea into a real product.
Comments