The Bluejay Times: First Edition

Hi. Hello. Welcome to the first ever Bluejay newsletter.

I'd like to set two goals for this newsletter. First, to give you an inside scoop of what's happening at Bluejay and the larger Voice / Conversational AI landscape. Second, and more importantly so, to give you a break from AI-generated consumption. When you read these, I want you to know that you are experiencing it from the source. Pure, human-to-human communication. I'm Faraz, Co-Founder and CTO here at Bluejay, and I'm honored to welcome you to The Bluejay Times: First Edition.

For those of you who are joining us for the first time (all of you 🙂), Bluejay is a testing and monitoring platform for Conversational AI agents. Companies ranging from Fortune 10 enterprises to fast-growth startups in the Silicon Valley use Bluejay to make sure their voice and text agents work in production (monitoring) and development (testing) environments. Our team, now seven strong, works around the clock to make sure your agent behaves when talking to customers.

If you are reading this, you fit into one of the following buckets:

1. You are a Bluejay customer (looking forward to seeing you at our next check-in!)
2. You are interested in Voice AI, Evals, or Voice AI Evals (welcome to the Bluejay family!)
3. You are an investor (Rohan just blew 50k in Vegas)

In any case, you can expect to receive weekly (a commitment has been made) newsletters that cover the week's happenings, and, to keep things real, an Author's Note heart-to-heart from a (currently) seed-stage founder navigating his way through the thrilling, magical, mind-bending world of SF startups.

Let's dive right in.

Announcements

Here's what happened at Bluejay last week:

• Rohan flew to San Diego and refined a Bluejay Sentiment Analysis dashboard for a customer by listening to every. single. call. They do 1M+ minutes a month.
• Faraz designed and implemented RBB, a novel technique for increasing determinism in LLM evals by Bluejay. This resulted in a 2x increase in eval determinism, and a 4.5x reduction in output variance – exciting stuff! An in-depth technical article will be released on this soon.
• The team released Bluejay's Frontend V2: the love child of beauty and speed.
• Last week, the team pushed 31190 lines of code with the sole purpose of bringing evals that work out-of-the-box to voice AI.

Feature Spotlight: Scripted Responses

Introducing Scripted Responses: an exciting new way to program Digital Humans to say exactly what you want them to say. Perfect for fine-tuning compliance, intent-understanding, and sending DTMF codes!

Author's Note

Bluejay is taking off. Flying, even. Yash (meet Yash!) spent the past few days optimizing our dashboard load times, as the queries he wrote to handle a few thousand calls no longer scaled for the millions of minutes of audio we recieve each month. How did we get here? By being diligent, fighting for every deal that we had (or didn't have), and by listening to our customers. To that end, please reach out if you have any feedback for Bluejay, or if you'd like to see any features be added to the platform. My email is faraz@getbluejay.ai!

Coming Soon

• Upcoming Events: Join Bluejay, Agora, and Assembly AI for a Voice AI Mixer on February 26th! Spots are limited!
• Upcoming Features: Revamped Alarms! With this feature release, customers will now be able to set time-based thresholds on their alarms so that they don't get notified each time their agent fails a custom metric evaluation. One failure could be a fluke, but 10 times over the past 5 minutes? That's important!

That's all for this week. I'll see you next time!

Faraz Siddiqi
Co-Founder & CTO @ Bluejay