Arjun Bhalla Email & Phone Number

Seattle, Wa, Us

Provo, Ut And Seattle, Wa, Us

· Maintain end-to-end data pipelines for data annotation projects from data handoff all the way to final delivery· Coordinate with data scientists and machine learning developers to develop and maintain best practices for formatting of datasets for easy ingestion into their models· Work with external data annotation platforms for upload and compiling annotated data for ML datasets, as well as make use of external APIs to gather statistics and gain important insights on all annotation projects running through the pipeline· Aid with technical difficulties on an ad hoc basis related to annotation platform functionality, and data handoff ingestion, and formatting; requests may come from annotation contractors, annotation project managers, or data scientists

Seattle, Washington, Us

· Manage project runs and deliverables for data collection projects, mainly text and speech for training data to be used for NLP algorithms; gather requirements from customers, create a plan of action, allocate resources, and communicate with other internal stakeholders· Manage the technical side of project execution; uploading crowdsourcing tasks to platform, periodically pulling completed data, and doing file manipulation to adhere to clients’ desired formatting· Monitor and track project milestones, and write scripts to create deliverables according to clients’ needs· Provide input on data quality requirements and best practices, and validate against customers’ quality requirements· Create and analyze data reports for internal review (usually done on a per-project basis; scripting to create statistical breakdowns for specific data projects if/when desired) · Regularly sync with “project squads” (group of employees from different teams assigned to a common project) to provide the scripting and technical guidance required for project execution, to ensure that business needs are met and data collections delivered on schedule.

Burlington, Ma, Us

Contract role, combining linguistic annotation and development work.Develop grammars and lexical resources, in the form of dictionaries, extensive training data labeled for intent at sentence level and named entity tagging at word level, and JSGF-formatted rule based grammars, to provide coverage for domain-specific commands in Nuance’s voice assistant software.Perform data staging and data builds with newly-annotated data, and use knowledge of machine learning and the QA process to examine accuracy rates and perform error analysis, and devise solutions accordingly.

Bellevue, Wa, Us

Worked on the language modeling and understanding portion of Voicebox's natural language dialogue systems, which primarily involved grammatical parameter-tuning for rule-based language comprehension models. Fixed comprehension errors resulting from rule-based models and tuned grammars effectively to a minimum accuracy of 96%.Contributed to R&D for newer statistical NLU models being developed by my team, including writing scripts for file processing and organization of utterance+intent combinations. Provided both scripted automation and manual annotation for tuning accuracies of fledgling statistical intent models.Coordinated with a team of NLU engineers, as well as data scientists housed within my team, for key NLU engine components such as Named Entity Recognition (NER) modeling and other machine learning techniques such as match quality for statistically determining intent of a given phrase, and classifying into a given conversational domain (music, weather, hands-free dialing, etc.). Also coordinated with English-language as well as foreign-language localization experts, and the engineers who design the conversational domains and databases from the ground up.

Bellevue, Wa, Us

Researched algorithms and approaches for a Natural Language Understanding (NLU) Plugin intended for various Voicebox products. Used both linguistic and computational analysis to identify issues and propose solutions in text interpretation for major European languages (German, French, Italian, Russian).Worked closely with several professional Computational Linguists, all of whom are housed within the NLU team, the department in which I interned. I learned of the day-to-day tasks of a Computational Linguist in a professional setting, as well as the close interaction required with Data Scientists (who make up a portion of the team).

Master’S Degree, Computational Linguistics

Master’S Degree, Linguistics

Bachelor’S Degree, Linguistics