Job Description

Qualifications


4 + years in data science or machine-learning roles (NLP or time-series a plus).


Hands-on with clustering, anomaly detection, and causal-inference techniques.


Strong Python (pandas, scikit-learn, PyTorch/TensorFlow) and solid SQL.


Experience turning notebooks into production jobs via Airflow, Flyte, or similar.


Comfortable with vector databases and similarity search (Faiss, PGVector, etc.).


You think in experiments, communicate results clearly, and love shipping incrementally.

Key Responsibilities


Pattern Discovery - Build unsupervised / semi-supervised models (HDBSCAN, metric

learning, spectral, etc.) that group similar issues and surface them for human review.


Trend Detection & Forecasting - Design change-point and anomaly detectors, then

forecast issue volume with Prophet, NeuralProphet, or your tool of choice.


Cross-Channel Correlation - Link signals across chat, email, voice, and social to

reveal how pain points bounce between channels.


Root-Cause & Prescriptive Modeling - Apply causal graphs or lightweight GNNs to

suggest likely drivers and remediation actions.


Active-Learning Loops - Create feedback workflows that let analysts merge/split

clusters and continuously improve model accuracy.


Experimentation & Metrics - Define success criteria, run controlled experiments, and

publish clear, visual results for the team.


Collaboration - Partner with the LLM, platform, and dashboard engineers to deliver

end-to-end features--then measure the lift.

Job Type: Full-time

Pay: Up to ₹1,000,000.00 per year

Benefits:

Provident Fund
Schedule:

Morning shift
Supplemental Pay:

Performance bonus
Experience:

machine learning: 4 years (Preferred) data science: 4 years (Preferred) python: 2 years (Required)
Work Location: In person