Startups & Projects

Here are the startups I've founded and the projects I've undertaken.

Inteliver AI SAAS

Inteliver AI SAAS

Modify and apply all kinds of transformations on your images on-the-fly and deliver customized tailored data to your users.

Learn more →
FarsAva AI

FarsAva AI

Revolutionized Persian speech recognition accuracy by inventing FarsAva, an advanced LSTM-based large vocabulary Automatic Speech Recognition system, using end-to-end Deep LSTMs and Distributed GPU training in PyTorch, surpassing all previous models by an impressive 14%.

Learn more →
zap.dev Agile Agency

zap.dev Agile Agency

Co-founded zap.dev, a dynamic and full-cycle web and mobile application development company specializing in creating tailored solutions for startups and fast-growing companies.

Learn more →
Cogneunce Startup

Cogneunce Startup

Cogneunce is a startup building the next generation of mental healthcare solution. Using Cogneunce mobile app patients can reserve an appointment with their therapist and hold their sessions online.

Learn more →
metierify Startup

metierify Startup

Metierify helps users to efforlessly create online professional resumes. CV imporevment suggestions from proffessional advisors along with AI-assisted resume writer makes Metierify one of the best resume builders.

Learn more →
Speech Segmentation System

Speech Segmentation System

Implemented a speech segmentation classifier that accurately discriminated between speech, non-speech, and silence, trained on a 300-hour dataset, utilizing SVM, and Scikit-Learn, achieving a 96% accuracy rate.

Learn more →
Designing and Implementing Farsava Speech Corpora Data Pipeline

Designing and Implementing Farsava Speech Corpora Data Pipeline

In this project, I took the lead in designing and implementing a data pipeline to create the Farsava Speech Corpora, a significant achievement that revolutionized the available Persian speech datasets.

Learn more →
Grapheme-2-Phoneme (G2P) Model for Persian Speech Recognition System

Grapheme-2-Phoneme (G2P) Model for Persian Speech Recognition System

As part of developing a Persian speech recognition system, we took charge of creating a crucial subsystem: the Grapheme-2-Phoneme (G2P) model for the Persian language. This subsystem was vital in generating a pronunciation dictionary for accurate speech recognition.

Learn more →