KabuK Style Inc.のMid-Level NLP Data Engineer - Price Comparisonの求人

KabuK Style is a travel tech company that provides the travel wallet service "HafH" with the mission of "Creating infrastructure for a society that embraces diverse values as they are."

As the first step toward building this world-changing infrastructure, we launched HafH in April 2019.

Currently, we have partnered with over 10,000 accommodation facilities across 130 countries and regions, with more than 120,000 members using our service.

We continue to expand as a platform that supports diverse lifestyles while revolutionizing the travel and hospitality industry, making travel easier, more convenient, and more affordable.

<Mission> Kabuk Style is a fast-growing fintech travel company operating HafH, an innovative OTA (Online Travel Agency).

To stay competitive, Kabuk Style continuously monitors hotel room prices offered by competing OTAs. However, each OTA uses different naming conventions and levels of granularity, and the underlying data is often unstructured. As a result, ensuring fair and accurate price comparisons is a significant technical challenge.

In this context, Kabuk Style is seeking a candidate with proven experience in data cleansing and NLP (Natural Language Processing) to tackle this problem.

The mission includes not only developing matching and comparison algorithms, but also ensuring their deployment on our cloud infrastructure, as well as continuously monitoring accuracy and making adjustments over time.

This is a fully remote position, preferably based in Japan. You will be fully dedicated to the project described above during the first 6 months. After that, we expect you to progressively contribute to other projects related to data, machine learning, and AI.

Because some of our data is in Japanese, knowledge of the Japanese language is welcome but not required.

<Why Choose This Position> - You will own a core technical project and have room to express your expertise and creativity. - Your work will have a direct impact on Kabuk Style’s core business by optimizing pricing and maintaining rate parity, improving conversion and revenue. - We offer flexible working hours and are primarily results-focused. - You will work with people from diverse cultural and professional backgrounds. - You will have opportunities to contribute to a variety of projects across Kabuk Style. - Kabuk Style is a fast-growing startup, which comes with risk but also significant growth potential.

---

<Application Process> Please submit the following documents: - Resume. - Short Cover Letter (optional) - If you choose to include one, please avoid generic or templated letters - Short report describing a real NLP project you have worked on in the past - Keep it concise: longer is not better (half a page to one page is ideal) - Choose a project as close as possible to the mission described above - Clearly describe: - the initial problem - the technical approach and solution - how accuracy or quality was measured - the business impact - the possible improvements or next steps - If some information cannot be disclosed, please explicitly mention it in the report

We are primarily looking for a mid-level candidate, typically with at least 4+ years of experience in AI and NLP. A research background (e.g. PhD or postdoc) is welcome, provided there is proven experience building and operating production systems.

We may also consider senior software engineers with limited prior experience in data or NLP, but with 10+ years of experience building large-scale software systems and a strong motivation to acquire new skills.

<Technical Skills> We expect you to have proven experience with applied NLP at scale: - Experience with the Python data analysis ecosystem: Pandas, Polars, scikit-learn, Matplotlib, Plotnine, etc. (Equivalent experience is also acceptable.) - Proven experience with modern NLP in Python: spaCy, NLTK, transformers, RAG, etc. (Equivalent experience is also acceptable.) - Solid understanding of the foundations of statistics and machine learning: regression, classification, clustering, recommendation systems, PCA, SVM, etc. - Strong understanding of cloud-based distributed computing. - Ability to write SQL queries involving JOINs, aggregations, and window functions. <Soft Skills and Culture> - Be independent: do not wait for your manager to assign tasks. - Be organized: keep task tracking up to date and maintain a clear view of priorities. - Communicate in English: all work is conducted in English, including meetings and chat. Your manager can also speak Japanese and French. - Collaborate actively with members of your team and other teams. - Be honest and transparent: mistakes are acceptable, and we do not expect you to know everything. - Provide constructive feedback: be direct yet professional. - “Disagree and commit”: once a decision is made, align with the team. - Stay motivated and productive in a fully remote environment.

<Technical Skills> - Experience with API-based AI systems from major providers such as OpenAI, Google (Gemini), Anthropic (Claude), or Microsoft. - Experience configuring serverless systems on Google Cloud or AWS. We have dedicated infrastructure engineers, but solid understanding is a plus. - Experience with web scraping using Python or Node.js. We have an engineer responsible for scraping, but the ability to understand and modify code is important. - Familiarity with the BigQuery dialect of SQL. <Soft Skills and Culture> - Ability to read Japanese, as some of our data is in Japanese. - A degree in computer science (Master’s level or higher). - Living in Japan, or in a country with at least 4 hours of working-time overlap with Japan (UTC+9).