The current app is being rebuilt as it sucks and the device is under active development.

We've gotten the following to work:

1. Emotion detection with barks

2. Cough detection (Kennel Cough specifically)

3. Identifying a dog from their bark

4. Video analysis of dog behavior

5. Identifying key parts in dog vocalizations (similar to phonemes in people)

6. Some basic intent detection (or what people call translation)

If anyone is good at μpython + TFLite and can help us transferring our models on device that'd be awesome. Currently our set up is super hacky.

Fullstack Hardware Engineer: Build smart collar hardware stack - PCBs, 3D printing, battery-efficient firmware. We have prototypes; need you to own hardware so we focus on distribution. Need: consumer product experience, 3D printing, firmware skills. Mostly C.

Canine Language Specialist: Ground AI models in real dog behavior. Design training protocols, collect data, teach dogs structured communication (buttons/vocals). Need: dog cognition experience, AAC systems, scientific mindset. Company: Decoding dog communication via smart collars using audio/motion/behavior data to understand what dogs are saying.

Apply: Email work samples, subject "{Dog's name} says {something}," explain what interspecies communication means to you.

Company: We’re building smart collars and devices to decode dog communication—using audio, motion, and behavior data. Our goal is to understand what our dogs are trying to say and rethink the worlds of training, healthcare, and companionship as we uncover their language. We've trained models for dog vocalizations and movement patterns, and now we’re integrating behavior and intent.

Role: You’ll work closely with our engineering team to help ground our models in real-world behavior. From structured training sessions to experimental communication techniques (e.g. dog buttons, symbolic vocalization training), your work will directly shape how our devices interpret canine communication. You'll guide the design of training protocols, data collection, and help teach dogs how to speak using structured methods.

You:

+ Experience in dog cognition, vocalization, and training techniques.

+ Familiar with AAC devices or button-based communication systems.

+ Scientific mindset with hands-on experience.

+ Passionate about language and cross-species understanding.

+ Must have a dog (ideally one you’ve trained using vocal cues or buttons)

Apply: Send us an email with a video or write-up of your past work (bonus if it includes your dog using buttons or vocal cues) to hello@sarama.ai with subject: “{Your dog’s name} wants {something they want}.” In the body, tell us what interspecies communication means to you.

---

If you refer someone to us, tell them to let us know and we'll send you a gift! (we aren't able to give a cash bonus just yet soon though)

Company: We’re building smart collars and other devices to translate dogs into our language—using audio and motion data. We want to know what our dogs want and are on a mission to rethink the entire healthcare & training industries as we better understand them. We are mostly software people and have built our own models for dogs (soon expanding to my new cats).

Role: Hardware, firmware, and software. Moving fast will be a mix of LLM, research, hacks, and rapidly typing. Build fast, no hand-holding. First product’s a warm-up. After that, it gets real.

You: Full-stack in the real sense (hardware + firmware + software) Fast, independent, opinionated Move fast, care deeply Must have a dog Culture: Obsessed with dogs. We’re building the interface for interspecies communication. Not just “pet people”— we believe dogs are family members fully including legally (we can discuss the changes in law on our call).

Apply: Send us an email with your portfolio (or something similar) with a picture of you and your dog to hello@sarama.ai with subject “{Your dog’s name} likes {something they like}” and in the body include what this tech means to you.

—-

Sarama | Cultural Savant | Remote | Full Time | Must have a dog | https://www.sarama.ai

Company: We’re building smart collars and other devices to translate dogs into our language—using audio and motion data. We want to know what our dogs want and are on a mission to rethink the entire healthcare & training industries as we better understand them. We are mostly software people and have built our own models for dogs (soon expanding to my new cats).

Role: This isn’t just marketing or vibes. You’ll shape product, brand, and behavior. You know how culture moves before it moves. You understand what dogs mean to us now—and what they could mean next.

You: Fluent in design, language, and social undercurrents Can shape narrative, product positioning, and human-dog interaction Unafraid to challenge Bernays view of the world Must have a dog. You live this. Culture: Obsessed with dogs. We’re building the interface for interspecies communication. Not just “pet people”— we believe dogs are family members fully including legally (we can discuss the changes in law on our call).

Apply: Send us an email with your portfolio (or something similar) with a picture of you and your dog to hello@sarama.ai with subject “{Your dog’s name} likes {something they like}” and in the body include how you see culture adapting.

Source: https://support.apple.com/guide/iphone/models-compatible-wit...

Sarama is the simplest way to implement multi-modal ML models in mobile apps, we're working with companies from pre-launch to unicorn businesses.

Apply here for Senior Swift Engineer: https://forms.gle/oJbQjJCfi46KyBhv5.

Apply here for Senior ML Engineer: https://forms.gle/QgLAB7VYZoqyQkPu6.

Apply here for ML Solutions Engineer: https://forms.gle/ivCXVfsiUiSNmBRL8.

Email any questions: praful@sarama.ai

Reference:

1. https://www.laika.berlin/en/blog/new-client-barkgpt-ai-dog-b...

1. https://www.smithsonianmag.com/smart-news/scientists-record-... 2. https://www.npr.org/2022/10/31/1132951238/what-sound-does-a-...

A few things we observed: 1. Dogs are more communicative once they believe you are correctly responsive to their vocalizations 2. Their barks are more differentiated over time and they start to introduce new types of vocalizations or sequences of vocalizations 3. Dogs don't have a structured vocal language as it starts of slightly differentiated and mostly to get people's attention

This isn't true for other animals (based on our reading of academic papers) and a commonality we've seen in the papers: as animals have more individuals they interact with across their life especially not directly related, their vocalizations are more structured. Within family units they tend to rely more on touch, body language, and some intermittent vocalizations. Though marine mammals are very vocal given they operate within acoustic environments in the ocean and it gets dark around 100m - 200m.

Basically, as number of sustained individuals in a group goes up and the number of non-related individuals increases, vocalizations get more diverse. We see cows, pigs, goats, and other herd animals are also very vocal which is how scientists have been decoding their speech recently similar to prairie dogs.

Much of what I've written here is a super simplification but happy to get deeper into the weeds.

Sarama | Bioacoustics Engineer | REMOTE (anywhere)

We're developing a vocal language between dogs & people using ML to analyze barks to decipher which ones convey meaning. We've spent the last few years understanding how to iterate on interspecies communication. We're looking for someone who deeply understands audio data and a huge plus if you've done work with non-human audio datasets.

Working on a new app design and new features to improve adding barks to our corpus.

praful@sarama.app

We're publicly launching our app with our new design.

praful@sarama.app

We're developing a vocal language between dogs & people using ML to analyze barks to decipher which ones convey meaning. We've spent the last few years understanding how to iterate on interspecies communication. We're looking for someone who deeply understands audio data and a huge plus if you've done work with non-human audio datasets.

Our website is still being developed but we have users actively using our alpha product and are rapidly iterating on our beta release.

praful@sarama.xyz