Ionic D.

Key Skills

Software

No software listed

Top Subject Matter

Aviation speech recognition (ASR) and forced alignment for noisy communications.

Multilingual speech translation and streaming speech recognition.

Dialogue/NLP for hospitality and clinical text summarization.

Top Data Types

Audio

Text

Image

Top Task Types

Transcription

Text Summarization

Freelancer Overview

Senior AI/ML Engineer — ASR data/training pipeline and transcription/forced-alignment target generation (Walmart Global . Brings 11+ years of professional experience across complex professional workflows, research, and quality-focused execution. Education includes Bachelor of Science, University of Illinois Urbana-Champaign (2016). AI-training focus includes data types such as Audio and Text and labeling workflows including Transcription, Translation, and Localization.

Labeling Experience

Senior AI/ML Engineer — ASR data/training pipeline and transcription/forced-alignment target generation (Walmart Global Tech).

AudioTranscription

Designed ASR training workflows for aviation audio using PyTorch/ESPnet with distributed GPU training to improve model learning and recognition quality. Built large-scale transcription and forced-alignment pipelines for noisy radio communications to generate high-quality training targets. Optimized domain vocabulary decoding and language model adaptation to improve transcription accuracy in real-world conditions. • Improved labeling efficiency for noisy radio communications. • Developed evaluation frameworks for WER/CER and diarization accuracy. • Deployed streaming ASR inference and preprocessing for cockpit audio robustness. • Collaborated with stakeholders on iterative model and data pipeline delivery.

2022 - 2026

AI/ML Engineer — training data preparation and model development for multilingual speech-to-speech translation (Nexus Commerce).

Audio

Developed a real-time, low-latency speech translation platform using streaming ASR and Transformer-based neural machine translation across 15+ languages. Created a domain-adapted speech-to-speech translation system for religious lectures and sermons, optimizing for dialect terminology and contextual accuracy. Built additional AI systems that likely required training data preparation and model supervision for production performance. • Engineered streaming ASR components for translation. • Specialized translation outputs for dialect and context accuracy. • Participated in end-to-end model development using deep learning pipelines. • Delivered real-time performance improvements for live meeting scenarios.

2018 - 2021

AI Software Engineer — NLP training/annotation-oriented work for clinical insights and summarization (TriHealth Data Sciences).

TextText Summarization

Deployed an AI concierge system for hospitality operations using voice and text interfaces, supporting autonomous handling of guest requests with dialogue management and early neural machine translation. Developed NLP capabilities for extracting insights from clinical notes and unstructured medical data to improve search and summarization in electronic health record environments. Work involved preparing and leveraging labeled or structured training inputs for dialogue and clinical NLP tasks to improve accuracy in production systems. • Improved search and summarization quality over unstructured medical text. • Supported entity-like extraction and summary generation for clinical notes. • Built conversational/translation-driven experiences requiring supervised NLP training. • Delivered improvements for operational autonomy and comprehension quality.

2016 - 2018

Education

Work History