Job Description

Title: NLP Research Engineer

About AI Research

The goal of J.P. Morgan AI Research is to explore and advance cutting-edge research in AI, including ML as well as related fields like Cryptography, to develop and discover principles of impact to J.P. Morgans clients and businesses. J.P. Morgan AI Research has assembled a team of experts in many fields of AI. They pursue primary research in areas related to our research pillars as well as concrete problems related to financial services. They partner with internal teams to accelerate the adoption of AI within the firm. They also work with leading academic faculty around the world on areas of mutual interest. The team is headquartered in New York and present in London and the Bay Area. Conducting AI research in financial services offers unique and exciting opportunities for impact -- as a member of this highly visible team, you will have the opportunity to realize significant impact not only within J.P. Morgan but also to the broader AI community.

NLP at AI Research

Textual corpora such as news articles, social media posts, enterprise documents, contrats, and similar documents compose a large part of datasets that J.P. Morgan employees interact with on a daily basis. As a team that aspires to transform the business using cutting-edge AI, we are deeply invested in research that discovers ways to efficiently process, transform, or generate such corpora. The team is currently exploring state-of-the-art research in areas such as question answering, natural language generation, knowledge representation and reasoning, narrative analysis, synthetic text data, and representation learning. Head over to our publications page to learn more about what we do.

Requirements:

A Master of Science or higher degree in Computer Science or related fields is required. As an NLP Research Engineer, we expect you to be fully familiar with:

- Python programming (highly proficient to expert level)

- Processing documents in xlm/html, pdf, json, docx, or other formats, using tools such as tesseract, pdfminer, beautifulsoup, openpyxl

- Web scraping and crawling tools such as python requests, selenium, phantomjs

- Search and indexing tools such as ElasticSearch or Lucene

- At least one web-dev framework such as Flask, REACT, AngularJS (preferred)

- Visualization tools such as D3, plotly, seaborn (D3 preferred)

- Text-processing tools such as Spacy, NLTK (Spacy preferred)

- NLP frameworks and libraries such as HuggingFace, gensim, stanza, fastText

- Common text processing techniques such as tokenization/lemmatization, part-of-speech tagging, chunking, segmentation, text similarity metrics, and regular expressions

- End-to-end development of data-heavy pipelines

- Deployment and management of cloud instances

It will be a big advantage if you:

- Have hands-on experience with diverse textual corpora such as news articles, enterprise documents, microblogs, conversational data, etc.

- Are familiar with financial documents such as SEC filings and market risk reports.

- Have worked on building, tuning and evaluating NLP models including text classification, topic modeling, and data extraction

- Have hands-on experience with big-data processing frameworks such as MR, Apache spark.

- Are familiar with containerization technologies

- Have worked with common graph databases and know basic graph processing and traversal algorithms

- Have some familiarity with CV tools and frameworks such as OpenCV, or with speech processing

But most importantly, we want you to be a resourceful problem solver and an enthusiastic learner.

Responsibilities:

As an NLP Research Engineer, you are expected to:

- Use internal systems, third-party APIs, or web-scraping tools to collect and pre-process textual corpora

- Build end-to-end pipelines for NLP applications, starting from document ingestion and pre-processing, to modeling, front-end dev and deployment

- Liaise with technology teams and provide support during the productization cycle

- Create advanced visualizations that facilitate user-friendly functionality and model explainability

- Conduct feature-engineering experiments and partner with researchers during scoping and feasibility studies

- Build analytical tools that provide real-time insights on data, models, and end-user engagement

- Provide mentorship and guidance to researchers on best practices related to development, engineering and deployment of NLP tools

In exchange, you will:

- Work closely with NLP researchers on highly impactful and transformative applications focused on cutting-edge NLU and NLG

- Liaise with expert technology teams and business partners across the Firm

- Have an opportunity to publish at top-tier conferences, journals, and other research venues

- Work on highly ambitious, high-visibility projects at the cutting-edge of NLP and AI research