Remote interviews are the new normal, with over 60% of HR professionals now using video during the recruitment process. Many companies are turning to hiring intelligence platforms like Screenloop to help manage this virtual interview process at scale.
Hiring intelligence platforms offer a variety of tools that help HR and management teams promote easy collaboration, facilitate candidate skill and job matching, summarize key sections of interviews, surface searchable insights, and reduce hiring bias.
The foundation of these tools? AI-powered transcription.
That’s why when two of Screenloop’s co-founders, Nuno Saldanha and Joao Leal, saw an opportunity to get into the hiring intelligence market, their first step was to embed AI into their platform that would help them provide automatic and highly accurate interview recording transcriptions to their customers.
For their customers, intelligent transcription would translate directly into increased productivity and decreased time-to-hire, as it would help cut down on time wasted by tedious manual tasks such as interview transcription, note-taking, and candidate review.
Nuno and Joao began looking for the best AI transcription model to meet their needs, evaluating accuracy, price, punctuation and casing, ability to transcribe heavily accented English, and availability of additional AI models.
The co-founders also wanted a partner that served as a conduit to new AI breakthroughs, so they could ensure the AI models they chose would continuously be state-of-the-art. This search led them to AssemblyAI.
“It was like night and day,” Nuno says. “AssemblyAI was the clear winner.”
As a leading API platform for state-of-the-art AI models, AssemblyAI offers an AI stack for spoken data that is easy for teams like Screenloop’s to integrate and scale. The Screenloop team was particularly interested in how AssemblyAI’s Core Transcription model, which transcribes audio and video files at high accuracy and outputs formatted, easily digestible transcripts, all with a single API call, would get them off the ground quickly.Learn about AssemblyAI’s new state-of-the-art speech recognition model, Conformer-2
In addition, Nuno and Joao also liked that AssemblyAI offers a suite of intelligence models that let product teams build high ROI features on top of the transcription data. And since the extremely large AI models are built on the latest AI breakthroughs, Screenloop’s co-founders were confident that AssemblyAI’s AI models would always be best-in-class, benefitting their growing customer base.
Transcript readability was also an important metric for Nuno, Joao, and their teams to consider when building with AI-powered transcription.
Transcript readability is an often overlooked metric when comparing transcription models and APIs. Many models, while technically accurate or with decent WER scores, return transcripts that are missing basic paragraph structure, punctuation, capitalization, and speaker labels.
These models output a block of text that looks something like this:
When i'm gone for a while but hes always supportive so that always takes a lot of stress off and lets me play and its a lot easier sure wow well back to your college and pro career i know you are a usc player and im sure that was an amazing team experience but a lot of college players dont go on to go pro even though theyre incredible players and the college level is very high its a shame that there isnt much more interest in college tennis but how do you talk about mindset shift from choosing tennis as a career versus like business or coding or something you know and then making that decision from usc to just go pro one of my goals when I was little was to always play professional i think maybe some people just want to go to college or get a scholarship and then end there but i knew i always wanted to continue my tennis
Instead, modern AI models like AssemblyAI’s Core Transcription model apply Automatic Casing and Punctuation, Paragraph Detection, and Speaker Diarization (speaker labels), to each audio or video file processed with their API. Together, these additions make transcripts much more readable at a glance and facilitate a better experience for end users.
Compare this transcript, run through AssemblyAI’s transcription model, to the one above:
<Speaker A> When I'm gone for a while, but he's always supportive, so that always takes a lot of stress off and lets me play, and it's a lot easier. <Speaker B> Sure. Wow. Well, back to your college and pro career. I know you are a USC player and I'm sure that was an amazing team experience but a lot of college players don't go on to go pro, even though they're incredible players and the college level is very high. It's a shame that there isn't much more interest in college tennis. But how do you talk about mindset shift from choosing tennis as a career versus, like, business or coding or something, you know, and then making that decision from USC to just go pro? <Speaker A> One of my goals when I was little was to always play professional. I think maybe some people just want to go to college or get a scholarship and then end there, but I knew I always wanted to continue my tennis.
Screenloop’s co-founders were also impressed with the AssemblyAI transcription model’s ability to transcribe non-native English speakers without sacrificing accuracy, even with the presence of strong accents. This ability was extremely important for Screenloop’s ability to offer a high-quality product for its wide international customer base.
After one month, the product team at Screenloop decided to roll out AssemblyAI's transcription model to its customer base—and they have been thrilled with the results.
AssemblyAI's exemplary support and transparency built trust and reliability, Nuno and Joao continue, citing each interaction as “key in establishing a relationship moving forward” and for continuing to grow their use of AssemblyAI’s AI models.
With the AssemblyAI-powered transcription data, Screenloop’s team built a highly competitive platform that offers interview intelligence, training, and shadowing tools for its end users, and with great results. On average, Screenloop’s customers now realize:
- 90% less time spent on manual hiring and interview tasks
- 20% reduced time-to-hire for open roles
- 60% less candidate drop-off during the interview process
- 50% less rejected offers for open roles
The automated tools help support their customers’ internal fairness and equity initiatives. This trend to incorporate “ethical AI” into hiring platforms can increase transparency in the hiring process and ensure a fairer, more equitable process for all involved.
What’s next for Screenloop? Screenloop’s product team is actively looking to expand its product roadmap to include intelligent features that facilitate better training for hiring teams, more exhaustive analytics, and making the hiring experience easier and more efficient for its current and potential customers.
All of these new features will be dependent on continually state-of-the-art, highly accurate AI-powered transcription.