Data Scientist

Responsibilities

Frame business problems across functions as supervised-learning tasks, then prepare enterprise data and build, tune, and validate end-to-end models. Explain findings to stakeholders, define business-aligned success metrics, and work with engineering teams to hand off production-ready models with monitoring and documentation.

Requirements

Requires 4–6 years of data science or machine-learning experience, production model delivery, strong classical machine-learning and statistical foundations, and expert Python and SQL skills. A quantitative degree and experience with enterprise data platforms, Spark, MLOps, and stakeholder communication are required; deep learning and generative AI experience are also sought, with LLM frameworks a strong plus.

Skills

• Classical Machine Learning
• Supervised Learning
• Python
• SQL
• Pandas
• NumPy
• Scikit-Learn
• Statsmodels
• PySpark
• Feature Engineering
• Statistical Analysis
• Time-Series Forecasting
• MLOps
• Deep Learning
• Generative AI
• Stakeholder Communication

Benefits

• Flexible Work Schedules
• Work From Home
• Paid Time Off
• Paid Holidays
• Professional Development and Training
• Competitive Salary
• Family Medical Insurance
• Personal Accident Insurance
• Health Awareness Programs
• Extended Maternity Leave
• Annual Performance Bonuses
• Referral Bonuses

About this role

Accellor is an AI-native services firm focused on delivering measurable business
outcomes through advanced AI, data, and engineering capabilities. We help global
innovators and Fortune 100 enterprises modernize their architectures, scale
operations, and operationalize intelligence across the enterprise.
Our core focus centers on transforming advanced AI from proofs-of-concept into
production-grade systems. Spanning AI solutions, robust data services,
enterprise applications, and product engineering — we tailor our offerings to
the distinct operational needs of industries such as healthcare, life sciences,
telecom, retail, financial services, and technology.
Operating as an extension of our clients' teams, we combine design thinking with
technology-agnostic architectures to ensure faster time-to-value and seamless
system interoperability. From initial strategy to full-scale deployment, we
build intelligent, connected ecosystems that drive sustained enterprise value
and redefine the future of transformation.

ROLE OVERVIEW

We are hiring a hands-on Data Scientist with 4–6 years of experience building
and shipping classical machine-learning models on enterprise business data. The
role sits in a central AI team at an enterprise client that serves the whole
business, so the problems change with demand: sales, marketing, pricing,
operations, procurement, HR, finance.
The core of the role is classical, hands-on data science: supervised learning -
classification, regression, forecasting - on structured, tabular data, taken
from business question to validated model to production handover. Generative-AI
and LLM skills are welcome on top of that foundation, as the team applies both
where each fit.

KEY RESPONSIBILITIES

•        Pick up problems from any function as the business needs them and frame
them as supervised-learning tasks — e.g. deal win/loss and churn classification,
price and margin prediction, demand and revenue forecasting, lead scoring and
customer segmentation for marketing, supplier risk and spend classification in
procurement, attrition prediction in HR, anomaly detection in operations.
•        Build, tune and validate models end to end: feature engineering, model
selection, cross-validation, hyperparameter tuning, calibration and error
analysis.
•        Work with structured enterprise data from CRM, ERP, product master and
finance systems — profiling, cleaning, joining and resolving data-quality gaps
before modelling.
•        Run exploratory analysis and statistical tests (hypothesis testing, A/B
analysis, causal reasoning) to back recommendations with evidence.
•        Explain model behaviour to business stakeholders using interpretability
techniques (SHAP, feature importance, partial dependence) and plain-language
insight.
•        Package models for production with engineering teams: reproducible
pipelines, versioned artefacts, batch or API scoring, monitoring for drift and
performance decay.
•        Define and track success metrics (precision/recall, AUC, MAPE, lift,
uplift) and the business KPIs they map to.
•        Document approach, assumptions and results; present findings to
leadership.

REQUIRED SKILLS AND EXPERIENCE

•        4–6 years as a data scientist or ML practitioner, with a few models
delivered into production use.
•        Strong grounding in classical ML: logistic regression, tree ensembles
(random forest, XGBoost / LightGBM / CatBoost), linear and regularised
regression, SVMs, clustering, and time-series forecasting (ARIMA, Prophet,
gradient-boosted forecasting).
•        Solid statistics: probability, distributions, hypothesis testing,
confidence intervals, bias/variance, handling class imbalance and leakage.
•        Expert Python for data science — pandas, NumPy, scikit-learn,
statsmodels — and strong SQL on large relational or lakehouse datasets.
•        Experience with at least one enterprise data platform: Databricks,
Microsoft Fabric, Snowflake or similar, including Spark / PySpark for large
data.
•        Feature engineering on messy, real-world business data; comfortable
with missing values, categorical encoding, time-based features and entity
resolution.
•        MLOps basics: experiment tracking (MLflow), model versioning, pipeline
scheduling, and monitoring in production.
•        Ability to translate a business question into a modelling plan and the
result back into a decision — strong stakeholder and communication skills.
•        Degree in a quantitative field (statistics, mathematics, computer
science, engineering, economics or similar).
•        Experience with deep learning and neural networks (PyTorch or
TensorFlow / Keras) — MLPs on tabular data, sequence models for time series, or
embeddings — and judgement on when they beat classical methods.
•        Working knowledge of LLMs, generative AI or agent frameworks, applied
alongside classical models — a strong plus. Hands-on with one or more of
Microsoft Agent Framework (MAF), LangChain, LangGraph, CrewAI or Semantic
Kernel.

PREFERRED QUALIFICATIONS

•        Experience modelling sales operations, pricing, quoting or supply-chain
data.
•        Pricing, margin or revenue modelling: price elasticity, win-probability
curves, discount optimisation, deal-impact analysis.
•        Exposure to one business function such as HR, operations or
procurement, with examples of models that changed a decision there.
•        Experience with Salesforce data (opportunities, quotes, accounts) as a
modelling source.
•        Familiarity with causal inference or uplift modelling.
•        Azure data and ML stack (Azure ML, Fabric, Databricks on Azure)

BENEFITS

Exciting Projects: We focus on industries like High-Tech, communication, media,
healthcare, retail and telecom. Our customer list is full of fantastic global
brands and leaders who love what we build for them.Collaborative Environment:
You Can expand your skills by collaborating with a diverse team of highly
talented pe...