- Series-A US startup working in the supply-chain visibility domain
- Challenging NLP tasks: NER / Entity Matching, Text summarization, Language agnostic sentence representations, others
- Working closely with the Core Data Science Team
- 5-8k USD gross + company options (more details below)
At Altana we've built the first AI Knowledge Graph of the global supply chain -- the world’s most comprehensive representation of global commerce activity. This data asset, composed of billions of records, covers more than 40% of global trade, corporate ownership registries in over 200 countries, the global movements of goods, illicit web activity, and more. We do work with the best data that money could buy on a market, but also with the proprietary data.
Whom we are looking for?
The Core Data Science Team team is looking for talented NLP engineer(s) / research scientist(s) who knows how to build, evaluate (and possibly deploy) effective and scalable solutions for:
- Text summarization:
Example: very noisy goods description in customs declaration entry - extract the main information about the product, exclude things that are rather irrelevant to product - color, company address, telephone, invoice number, etc.;
- NER / Entity Matching:
Company names/types/details/address, product units/quantity/price/tariff code from customs declarations. We do currently use Prodigy for labeling and Spacy for model training - it produces quite good results but we need to push it further;
- Neural Machine Translation / Language-agnostic sentence representations:
We work with data in multiple scripts (English, Spanish, Portuguese, Arab, Russian, and many others). We need someone who will research and build a language-agnostic sentence representations (e.g. using Facebook's LASER or MUSE libraries, or propose something different that works)
The candidate should be proficient in English.
Tech stack (related to position):
- Pytorch / Tensorflow / Keras
- BERT (Roberta, Albert), Elmo, Fasttext, Spacy
Tech stack (used in company; not an exhaustive list):
- AWS, Azure
- Docker, Git, Kubernetes, Airflow, Spark, Swagger/OpenAPI
- Postgres, Amazon Redshift, Azure Synapse, Neo4j Graph DB
- ElasticSearch, FAISS
What we propose?
- Compensation: 5-8k USD gross, depending on the level of expertise
- Company options (standard vesting procedure)
- Compensation of the office-related expenses (you can join our teammates in co-workings of Kyiv / Lviv, or work entirely remotely from home)
- Family policy: paid parental leave 2 months for the primary caregiver and 1 month for the secondary caregiver
- 21 calendar days of paid vacation per year (15 business days)
- Support with PE account (FOP) - upon request
This position is remote, but you should be comfortable working as close as possible to a New York time - as a reference, our local team in UA works normally from 11.00 - 13.00 till 21.00 - 23.00.
About Altana AI
Altana AI provides a shared artificial intelligence model of the global supply chain to help governments, enterprises, and financial institutions to see across borders, manage risk, and improve global commerce.
About our company:
Job posted on
26 April 2021