Back to jobs

Senior Data Scientist, Applied ML

Pay
Pay not listedSimilar roles state $120K–$194K
Work setting
Not stated
Employment
Full-time — employment source
Employment type in the employer’s posting: Full-time
From the employer’s posting

What you’ll work on

Full posting
  • Scope projects with traders and analysts, each with a commercial hypothesis and a go/no-go point

  • Build LLM, NLP and multimodal pipelines that extract signals from outage reports, news and filings

  • Build models that predict outage impacts and detect anomalies and regime changes

  • Deploy proven models to production with CI/CD, monitoring and drift detection

  • Run quick feasibility experiments and stop weak ideas early

From the employer’s posting
turn market questions into scoped projects, each with a clear commercial hypothesis and a defined go/no-go point
Build LLM, NLP and multimodal pipelines that extract structured signals from unstructured sources such as outage reports, news, filings, and web content
Build models that predict outage impacts, classify disruption patterns, and detect anomalies and regime changes in fundamental data
Take the models that prove out into production. That means packaged, tested, deployed through CI/CD, with monitoring and drift detection built in from day one.
Design and run quick experiments to test feasibility across modelling approaches, and stop weak ideas early.

Before you apply

Employer’s visa record
10 H-1B petitions approved in fiscal 2024 and 2025For the employer as a whole. This posting doesn’t say whether the role is sponsored. US Citizenship and Immigration Services.

What you’ll bring

All qualifications

Core experience

  • Python and ML stack: Python with scikit-learn, PyTorch, and HuggingFace.
  • Communication: You can explain what a model means for a trade in two sentences, and you are comfortable being challenged by people who know the market better than you do.
Qualification wording
Python and ML stack: Python with scikit-learn, PyTorch, and HuggingFace. You write production-quality code that others can review and maintain.
Communication: You can explain what a model means for a trade in two sentences, and you are comfortable being challenged by people who know the market better than you do.
Education & alternatives
- Desk-facing track record: You have worked directly with commercial users in trading, finance, or a similar fast-paced setting, and you can point to decisions your work changed. - Quantitative foundation: Master’s or PhD in statistics, mathematics, physics, engineering, or equivalent depth gained in industry. Strong statistics and experimental design. - LLMs and NLP: Hands-on experience building extraction and classification pipelines with LLMs, covering evaluation, prompt and retrieval design, and cost and latency trade-offs.
Apply at Vitol

Posting record

Report an error

Actively hiringFirst seen by us 8 days ago, still listed, and not a repost of an earlier posting.

Status in our records
Live
First seen by us
Nov 13, 20268 days ago
Recorded sightings
9
Last seen by us
Nov 21, 2026

These dates show when we found the posting. Check the employer’s website to confirm it is still accepting applications.

How this job compares

Pay

$120K–$194K is typical

The middle half of what similar US postings state. This one doesn’t list pay.

From 3,983 postings in the last year. Stated ranges, not salaries paid.

Time since we first saw it

Day 8

Similar jobs usually come down between day 7 and day 60.

At Vitol, jobs usually come down by about day 33.

From 12,342 similar postings that came down, and 15 of Vitol’s. Coming down is not always a hire.

Vitol is hiring

4 open jobs

Open jobs by the week we first saw them, last 12 weeks.

  • Hiring is picking up
  • Takes jobs down at a usual pace
Hiring record

See how this role fits your experience

Add your resume and we’ll mark which of these tools you already use, and point out anything in the posting that could rule you out.

  • Python
  • AWS
  • Docker
  • Hugging Face
  • PyTorch
  • scikit-learn
Where the posting names them
- Take the models that prove out into production. That means packaged, tested, deployed through CI/CD, with monitoring and drift detection built in from day one. - Deliver insights where traders already work: Teams, alerts, and the Vision Excel and Python tooling. Standalone dashboards are not the default. - Source data through the Vision platform. When you need a new source, work with the Vision team rather than building side pipelines, and follow the existing web-scraping architecture.
- LLMs and NLP: Hands-on experience building extraction and classification pipelines with LLMs, covering evaluation, prompt and retrieval design, and cost and latency trade-offs. - Python and ML stack: Python with scikit-learn, PyTorch, and HuggingFace. You write production-quality code that others can review and maintain. - Production engineering: Docker, AWS, CI/CD, and MLOps practice, including model versioning, monitoring, and retraining.
- Python and ML stack: Python with scikit-learn, PyTorch, and HuggingFace. You write production-quality code that others can review and maintain. - Production engineering: Docker, AWS, CI/CD, and MLOps practice, including model versioning, monitoring, and retraining. - Time series and signals: Experience with anomaly detection and with combining signals across multiple time-series and text sources.

Your next step

Complete your application on jobs.smartrecruiters.com. The employer’s form will show what is required.

Applying through SmartRecruiters

  • Uploading your resume fills in the application form for you, so check what it filled before you send.

Already applied? Track this application

Job description

View original posting ↗

Job Description

Vitol’s Applied ML team in Houston works directly with the trading desks to find commercial value through data and applied data science . Projects start with a trader’s or analyst’s question, not with a model.

You will work alongside traders and analysts from the first conversation through to production and beyond. You frame the problem with them, review results with them as the work progresses, and stay with the tool after it ships.

You own both the science and its path to production. The work is fast and phased, and it is judged on one thing: whether the desk uses it to make better decisions.

What You Will Do

  • Work side by side with traders and analysts to turn market questions into scoped projects, each with a clear commercial hypothesis and a defined go/no-go point.
  • Review progress with the desk throughout each project, not just at handoff. Show intermediate results early, test them against the desk’s market knowledge, and change course when the signal or the priority shifts.
  • Build LLM, NLP and multimodal pipelines that extract structured signals from unstructured sources such as outage reports, news, filings, and web content.
  • Design and run quick experiments to test feasibility across modelling approaches, and stop weak ideas early.
  • Build models that predict outage impacts, classify disruption patterns, and detect anomalies and regime changes in fundamental data.
  • Take the models that prove out into production. That means packaged, tested, deployed through CI/CD, with monitoring and drift detection built in from day one.
  • Deliver insights where traders already work: Teams, alerts, and the Vision Excel and Python tooling. Standalone dashboards are not the default.
  • Source data through the Vision platform. When you need a new source, work with the Vision team rather than building side pipelines, and follow the existing web-scraping architecture.
  • Evaluate commercial AI platforms and tools against real desk problems, and give a clear recommendation.
  • Explain results, uncertainty, and limits in plain terms to traders, analysts, and senior stakeholders.

 

Qualifications

  • Applied experience: 5+ years applying machine learning and NLP to messy, real-world data, including models you took to production and supported after launch.
  • Desk-facing track record: You have worked directly with commercial users in trading, finance, or a similar fast-paced setting, and you can point to decisions your work changed.
  • Quantitative foundation: Master’s or PhD in statistics, mathematics, physics, engineering, or equivalent depth gained in industry. Strong statistics and experimental design.
  • LLMs and NLP: Hands-on experience building extraction and classification pipelines with LLMs, covering evaluation, prompt and retrieval design, and cost and latency trade-offs.
  • Python and ML stack: Python with scikit-learn, PyTorch, and HuggingFace. You write production-quality code that others can review and maintain.
  • Production engineering: Docker, AWS, CI/CD, and MLOps practice, including model versioning, monitoring, and retraining.
  • Time series and signals: Experience with anomaly detection and with combining signals across multiple time-series and text sources.
  • Communication: You can explain what a model means for a trade in two sentences, and you are comfortable being challenged by people who know the market better than you do.
  • Nice to have: Knowledge of energy or commodities markets, graph modelling, sentiment analysis, or agentic workflows and MCP.

 

Additional Information

What Success Looks Like in Year One

  • Traders and analysts on at least two desks bring problems to you directly.
  • At least two proof-of-value projects reach a go/no-go decision, and at least one is in production and in regular use on the desk.
  • Your production models are monitored, documented, and supportable by the wider team. Failures are found by monitoring, not by a trader.
  • Failed experiments are stopped early and written up, so the team does not repeat them.
  • You have built reusable components, such as extraction pipelines and evaluation harnesses, that the rest of the team uses.

The Environment

Vitol is one of the world’s largest independent energy traders, and Houston is one of its core trading offices. The Applied ML team sits close to the desks, so feedback is fast and the people using your work are just a few desks away.

We look for scientists and engineers who own outcomes, not just tickets. You will have room to choose your approach, and you will be expected to deliver work that traders can rely on.

Comprehensive benefit coverage includes:

  • Medical
  • Dental
  • Vision
  • Paid Vacation
  • 401k with company contributions
  • Life insurance
  • Short-term & Long-term disability

All your information will be kept confidential according to EEO guidelines.

Company Description

Vitol is a leader in energy and commodities. Vitol produces, manages and delivers energy and commodities to consumers and industry worldwide. In addition to its primary business of trading, Vitol is invested in infrastructure globally, with $10+billion invested in long-term assets.

Vitol’s customers include national oil companies, multinationals, leading industrial companies and utilities. Founded in Rotterdam in 1966, today Vitol serves its customers from more than 40 offices worldwide. Revenues in 2025 were $343bn.

Our people are our business. Talent is precious to us and we create an environment in which individuals can reach their full potential, unhindered by hierarchy. Our team comprises more than 65+ nationalities and we are committed to developing and sustaining a diverse work force. 

 Learn more about us here. This role is located in Houston, TX onsite 5x a week

 

Demand for Data Scientist roles

New Data Scientist postings at the employers we follow are up 3% over the last 6 weeks compared with the 6 before.

Weekly new Data Scientist postings at the same employers, from the week of August 17 to the week of November 9. The fewest in a week was 1,389, the most 2,932, and the latest week had 1,705.
August 17Last week

Each bar is one week of new postings. The darkest bars are the last 6 weeks.

Counted from the hiring pages we’ve followed for at least five weeks, so employers we add don’t look like growth. Updated weekly.

About the role

We file this job under Data Scientist, by the work the posting describes.

Data scientists use statistics and models to answer harder questions, such as whether a change worked, what demand will look like next quarter, or which customers are likely to leave.

Postings usually ask for Python or R, SQL, statistics, experiment design and common machine learning methods.

Unlike an ML engineer, a data scientist's main output is often an analysis or a first model, not software running inside a product.

US government figures for data scientists

Median yearly pay, May 2025
$120,230
Expected change, 2025 to 2035
34.6% more jobsagainst 3.5% for all jobs

Source: US Bureau of Labor Statistics, Occupational Employment and Wage Statistics (May 2025) and Employment Projections (2025 to 2035). Occupation code 15-2051.

Source details

View original posting ↗

Source notes

Source excerpts

Selected passages from the saved posting. Check the full description for conditions and exceptions.

Pay

No pay amount identified in the saved description.

Location & working pattern

Houston, TX, United States

Working pattern and location restrictions need checking in the full posting.

Work authorization

No clear work-authorization passage found. Eligibility is unconfirmed.

Find answers in the posting

AI
How answers work

AI selects complete passages from this posting. Check them for conditions and exceptions.

Uses this posting and your question. No profile needed.