Early level Data Scientist / ML Engineer

Remote • Posted 4 hours ago • Updated 4 hours ago
Full Time
No Travel Required
Remote
Depends on Experience
Fitment

Dice Job Match Score™

🔢 Crunching numbers...

Job Details

Skills

  • Machine Learning (ML)
  • Data Science
  • Git
  • SQL
  • Artificial Intelligence
  • LLM
  • Python
  • Apache Velocity

Summary

Level: Early career (1–3 years of project experience) with strong academic background in DS & ML.

About the role

We are building an Intelligence Platform: a platform that lets banks run AI agents safely.  You will build the data science and ML models that make bank controls work. We want someone early in their career who is curious, builds fast, and brings fresh ideas, not someone who has seen it all.

What you will work on

•         Data and labels. Profile alert, customer and transaction data; build the golden set of labelled past alerts with analysts; write clear labelling guidelines and measure how well labelers agree.

•         Decision modelling. Turn the bank's thresholds and risk tiers into a decision table; calibrate model confidence so it can be trusted; tune thresholds by weighing the cost of a wrong decision.

•         Features. Build transaction-behavior features (velocity, exposure) and help resolve records into a clean entity graph.

•         Safety models. Build and evaluate a detector for banking personal data (account numbers, IBANs), a grounding check that agent reasoning cites real evidence, and a prompt-injection classifier.

•         Retrieval and prompts. Choose embedding and reranking models; shape prompts and structured outputs for the triage agents.

•         Evaluation and model risk. Run evals in CI, test for fairness and stability, compare the agent against today's manual process, and write model cards that reviewers can rely on.

•         Later phases. Smart request routing, drift monitoring, anomaly detection on agent behavior, and fine-tuning small open-weight models on approved traces.

What we are looking for

•         1–3 years in data science or ML engineering, or a strong degree, thesis or portfolio that shows the same

•         Solid Python (pandas, scikit-learn) and SQL; writes clean, tested code and uses Git

•         Hands-on work with LLMs: prompting, structured outputs, embeddings or retrieval, and evaluating results

•         Sound grasp of classification metrics, calibration, train/test splits and data leakage

•         Comfort turning a vague problem into a small experiment, then explaining the result in plain words

•         A habit of proposing better ways to do things, and the honesty to say when an idea did not work

•         Experience with NER or text classification (Presidio, spaCy, Hugging Face)

•         Familiarity with MLflow, Feast, LangGraph, eval tools (promptfoo, DeepEval, Ragas) or fairness tools (Fairlearn)

Employers have access to artificial intelligence language tools (“AI”) that help generate and enhance job descriptions and AI may have been used to create this description. The position description has been reviewed for accuracy and Dice believes it to correctly reflect the job opportunity.
  • Dice Id: marlabnj
  • Position Id: 9144928
  • Posted 4 hours ago
Contact the job poster
Bharath Kumar

Bharath Kumar

Director -Recruitment @ Marlabs LLC

Create job alert
Never miss an opportunity! Create an alert based on the job you applied for.