Text and Speech Analyst

Arbitration Forums Inc.•Tampa, FL

16d•Remote

About The Position

This role at Arbitration Forums is as unique as it is rewarding because of the AF IPAAL Values (Integrity, Passion, Accountability, Achievement, Leadership) and TRI Model (Trust, Respect, Inclusion). The Text and Speech Analyst is responsible for analyzing, interpreting, and evaluating language and text data to facilitate effective utilization across various contexts in support of our AI-powered solutions. This role examines written, spoken, and digital communications to identify patterns, trends, and opportunities related to language and speech to benefit our members in a cost-effective manner. The Text and Speech Analyst collaborates with cross-functional teams in the design, development, testing, and deployment of text and speech solutions that produce accurate, relevant, and high-quality outputs across various business applications. The ideal candidate will have a strong background in AI, natural language processing, and neural networks.

Requirements

Bachelor’s or Master’s degree in Computer Science, Information Systems, Data Science, Linguistics, or a related field.
Minimum of 6 years of experience in NLP, LLMs, ASR engines and frameworks, data governance, data science, or a related role.
Working knowledge of ASR engines using frameworks like Wav2vec or Deep Speech.
Working knowledge of Azure AI Services.
Advanced programming knowledge, including mastery of programming languages such as Python, C++, Java, and especially AI-centric libraries like TensorFlow, PyTorch, and Keras.
Cloud computing and knowledge for deploying and managing AI applications on cloud platforms like AWS, Google Cloud, or Microsoft Azure.
Expertise in classical speech processing methodologies like hidden Markov models (HMMs), Gaussian mixture models (GMMs), Artificial neural networks (ANNs), Language modeling, etc.
Functional knowledge of Snowflake and/or Databricks for AI capabilities and data management.
Hands on experience on current deep learning (DL) techniques like Convolutional neural networks (CNNs), connectionist temporal classification (CTC) used for speech processing.
Working knowledge of techniques for text parsing, sentiment analysis, and the use of transformers like GPT (generative pre-trained transformer) models.
Exposure to open-source models such as Kaldi or OpenAI Whisper.
Data management knowledge, including data pre-processing, augmentation, and generation of synthetic data, including the cleaning, labeling, and augmenting of data to train and improve AI models.
Proficiency in Transformer Models like BERT-base. ELMo, ULM-FIT.
Strong understanding of natural language processing concepts and techniques.
Working knowledge of cloud services (i.e., MS Azure, AWS, Google Cloud).
Experience with AI tools, such as MS Azure ML, Databricks AI, Snowflake CortexAI, Dataiku.
Strong knowledge of data governance, data security, and compliance practices.
Familiarity with data visualization and reporting tools (e.g., Webfocus, Power BI).
Hands on experience with MLOps processes.
Excellent analytical and problem-solving abilities.
Strong communication and interpersonal skills to collaborate with cross-functional teams.
Ability to lead projects and mentor junior staff.

Nice To Haves

Auto Insurance claims industry experience preferred.

Responsibilities

Design, develop, and refine automated speech recognition models and contextual text mapping techniques to ensure high-quality outputs for specific use cases.
Employ techniques to guide and enhance Automated Speech Response and text harvesting models, ensuring that the AI interactions are effective and efficient.
Develop effective AI text and speech services through proficient programming and utilization or ASR engines using standard and fit-for-purpose frameworks.
Collaborate with cross-functional teams to ensure that the models are aligned with the business requirements and objectives.
Conduct testing of variants and analyzing model behaviors, including deviations and degradations within given frameworks.
Stay up to date with advancements in speech recognition and text harvesting technology and engineering best practices.
Collaborate with product managers, data scientists, ML engineers, and business stakeholders to integrate AI models into business processes, products, or workflows.
Create documentation and reusable libraries for internal and external use.
Partner with the MLOps Engineer and other stakeholders to establish and implement observability and monitoring frameworks to adequately and timely identify degradations and potential ethical/bias issues.
Establish and implement explainability frameworks.
Provide recommendations for improvement in the areas of AI acceptable use and ethics, collaborating with Legal and Compliance to ensure adherence.
Ensure data quality and integrity as they apply to text and speech models.
Ensure that data security protocols are followed in the definition and implementation of AI solutions in accordance with regulatory requirements and company policies.
Develop safety filters and guide the ethical use of prompts to prevent biased or harmful responses.
Work closely with IT, products, data engineers, data analysts, data scientists, and business stakeholders to understand needs, data requirements, and implement solutions.
Provide technical leadership and mentorship to junior data team members.
Provide training and support to team members on effective frameworks and best practices.
Stay updated on emerging technologies and best practices, making recommendations for continuous improvement.
Support model observability efforts to ensure adherence to company policies and enforce governance standards.
Other duties as assigned by manager or project needs