BoloSuno

🚀 Founding AI/ML Engineer (Co-Founder) - Bharat’s Audio Foundation Model

📌 Apply here: https://forms.gle/suLFDWfFEshCha7G7

 

About Us:

We’re building Bharat’s first at scale audio foundation models... transformer architectures trained entirely from scratch.

Think LLMs, but for audio understanding & generation.. no fine tuning, no shortcuts, 100% original models built for Bharat’s 150+ crore voices.

 

The Role:

You’ll architect, train, and deploy transformer based audio foundation models from the ground up.

From research to real time deployment, you’ll have end-to-end ownership of model design, infrastructure, and performance.

 

What You’ll Build:

* Audio Transformers: Encoder–decoder & decoder only models, attention optimized for sequential audio.
* Foundation Training: 100M+ parameter models on large scale Bharat audio datasets with self-supervised learning.
* Distributed Infrastructure: Multi-GPU/TPU pipelines with model parallelism, mixed precision, gradient checkpointing.
* Low-Latency Serving: Optimized inference systems with quantization, caching, and deployment pipelines.

 

Must Have Skills:

* Proven experience implementing transformers from scratch (not just using library defaults).
* Trained large scale audio models (speech/music/environmental) from first principles.
* Distributed training expertise for multi-GPU/TPU systems.
* Strong audio signal processing background (spectrograms, mel-frequency, feature extraction).
* Expert in PyTorch or JAX with custom architecture experience.

 

Nice-to-Have:

* Self-supervised learning for audio (contrastive learning, masked prediction).
* Speech AI (ASR, TTS, audio generation).
* Production deployment of large models.
* Published research in transformers or foundation models.

 

Why Join Us:

🛠 Early Builder - Shape Bharat’s AI led audio ecosystem.

📈 Equity Ownership - Significant stake, not just salary.

🎯 Creative Freedom - Architect without bureaucracy.

🌏 Bharat Scale Impact - Build for 22+ languages & dialects.

🚀 Visibility - Conferences, AI community recognition.

❤️ Mission Driven - Preserve and empower Bharat’s voices.

 

Let’s Build Bharat’s Audio Future

📌 Apply here: https://forms.gle/suLFDWfFEshCha7G7

Machine learning

Infrastructure

Research

Data

Community management

PyTorch

Audio Development

Unsupervised Learning

GPGPU

Low latency

Cache

Signal Processing

AI Engineer - Machine Learning | Artificial Intelligence | ML | NLP | Audio | Data Scientist | Transformer Foundation Models (From Scratch)