Linguist III

Job Title: Linguist III
Duration: 04 Months (Possible extension)
Location: Remote role. West coast preferred (cannot be located in TX or Illinois as those 2 states have passed laws that are incompatible with our data collection agreements)
Summary:

• Must have a PhD in Linguistics IN HAND. Not just working to get one.
• A graduate degree in Literature or English is not an appropriate substitution. A degree in Computer Science with a specialization in NLP is not an appropriate substitution. A degree in Computational Linguistics is not the same as theoretical Linguistics.
• Must have near-native or native bilingual proficiency in a language from the Dravidian family and professional fluency from the Indo-Aryan family. (Candidates that don't meet the SPECIFIC language requirement will be disqualified. We're looking for a specific language family here). Dravidian is more of a focus here and being fluent in both 1 Dravidian and 1 Indo-Aryan language is the best profile. Please make sure the resume is CLEAR on which language the candidate is being submitted with and the CEFRL level.

Top 3 must-have HARD skills:  

• Perform linguistic error analysis of machine translations and identifying the most frequent and severe error categories
• Strong skills in pattern recognition, cross-functional communication, and multitasking
• Experience with Python

Good to have skills:     

• Experience with creating and/or maintaining specialized lexical resources (e.g., profanity dictionaries) a plus
• Ability to independently work through ambiguous requests, based on priorities established by CWAM, and perform under pressure. Able to work cross functionally.

Main duties:

• Perform linguistic analyses on large datasets.
• Perform linguistic error analysis of AI model outputs, determining what the most frequent and severe error categories are.
• Write and revise guidelines for human annotation and other AI projects, including but not limited to translation tasks.
• Conduct typological and sociolinguistic research on a large number of languages, highlighting their similarities and differences.
• Perform linguistic analyses for Responsible AI (toxic language, hate speech, gender bias and other cultural biases) in massively multilingual settings.
• Conduct linguistic literature reviews on various NLP-adjacent topics, and summarize findings.
• Compare the quality of deliveries between vendors, identify error patterns, and provide actionable feedback.
• Provide information or guidance relative to any aspect of linguistic knowledge (typology, morpho-syntax, sociolinguistics, classification, phonetics/phonology, pragmatics, etc.).
• Reach out to and collaborate with native speakers in various languages.
• Communicate results of linguistic analyses to engineers and research scientists.

Skills:

• Must have strong written and spoken communication skills, especially business and research communication.
• Must have near-native or native bilingual proficiency in a language other than English, more specifically a language of the following families or groups: Dravidian or Indo-Aryan.
• Working knowledge in other languages is a plus. Proficiency in a low-resource language is valued.
• Must be able to code in Python (must) and query databases using SQL, other coding languages used for data analysis are a plus.
• Must be able to independently work through complex requests and perform under pressure.
• Strong ability to work independently, prioritize, plan, and track work, as well as report progress education or training in the basics of project management is a plus self-motivation is a must
• Working knowledge of international language-classification standards is valued.

Education:

• Graduate degree in Linguistics or related field is a must; PhD is a plus a background or specialization in corpus linguistics is a plus
• experience with field work is a plus
• Must have a very firm grasp of the following linguistic fields: language typology, syntax, morphology, sociolinguistics (especially dialectology and discourse analysis), corpus linguistics, writing systems, pragmatics, phonology.
• Must have some experience with applying basic Natural Language Processing techniques.

Experience:

• Years of experience: 0-3
• Experience working cross-functionally
• Experience collaborating with machine learning, NLP, or software engineers, or data scientists
• Experience contributing to research papers
• Important: Preferably no known conflicts of interest in the fields of machine translation, ASR, TTS, or LLM research (as FAIR Linguists need to be contributing to research papers)

Compelling Story & Candidate Value Proposition:
Here are some examples of the cool big things the FAIR C&L Linguistics team has provided impactful support for:

• NLLB (pivotless text-based translation system for 200 languages), now UNESCO's Universal Translator (https://huggingface.co/spaces/UNESCO/nllb ) and one of the top translation engines on Wikipedia for low-resource languages. The research was published in Nature (https://www.nature.com/articles/s41586-024-07335-x).
• MMS (ASR and TTS for over 1,000 languages), now UNESCO's ASR system (https://huggingface.co/spaces/UNESCO/MMS ) in support of the Decade of Indigenous Languages, as well as, more recently, Omnilingual ASR (https://ai.meta.com/blog/omnilingual-asr-advancing-automatic-speech-recognition/ ), which supports over 1,600 languages and enables simple zero-shot additions.
• Seamless (pivotless speech-based translation system for 100 languages), which was recognized as one of Time's best 2023 AI inventions (https://time.com/collections/best-inventions-2023/6326994/meta-seamlessm4t/ ). The research was recently published in Nature (https://www.nature.com/articles/s41586-024-08359-z#group-1 ).

 

At SPECTRAFORCE, we are committed to maintaining a workplace that ensures fair compensation and wage transparency in adherence with all applicable state and local laws. This position's pay range is $50.00/hr – $55.00/hr.