Data Scientist

Science Logic
Hyderabad
Workplace: HybridFull timeFunction: Data Science & Machine LearningExperience: 3+ yearsEducation: bachelorsSkills: ["Analysis","Communication","Experimentation","Problem-solving"]

Build and run evaluation and quality measurement for a production suite of small, locally hosted language models and agents. Define “good” for non-deterministic behavior, create regression and golden-set harnesses, and analyze response quality, retrieval performance, and robustness under adversarial conditions. Also develop production forecasting, anomaly detection, and early-warning models on operational telemetry, turning insights into recommendations for engineering and product within enterprise security constraints.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Science Logic
Science Logic
5 days ago

Data Scientist

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 20 hours agoStatus: Live

Job Summary

Build and run evaluation and quality measurement for a production suite of small, locally hosted language models and agents. Define “good” for non-deterministic behavior, create regression and golden-set harnesses, and analyze response quality, retrieval performance, and robustness under adversarial conditions. Also develop production forecasting, anomaly detection, and early-warning models on operational telemetry, turning insights into recommendations for engineering and product within enterprise security constraints.
Location: Hyderabad
Workplace: Hybrid
Employment Type: Full time
Job Function: Data Science & Machine Learning
Seniority: Mid level

Key Responsibilities

  • •Design and own evaluation harnesses for LLM and agent outputs, including golden sets, regression suites, and rubric scoring.
  • •Build and validate LLM-as-judge pipelines, calibrate judges against human labels, and track response-quality metrics like groundedness and hallucination rate.
  • •Conduct adversarial and robustness testing (prompt injection, jailbreaks, tool misuse) and create chaos/stress tests to expose failure modes.
  • •Evaluate retrieval and multi-step agent trajectories (recall@k, MRR/nDCG, tool-call correctness, replayable-state inspection) and assess routing quality.
  • •Build production predictive models from operational telemetry (forecasting, anomaly prediction, early-warning signals) and wire signals into the LLM/agent layer for recommendations.

Key Requirements

  • •3+ years in data science, ML, or applied quantitative analysis.
  • •Strong applied statistics and experiment/significance-test judgment for noisy, non-deterministic outputs.
  • •Experience building, deploying, and monitoring predictive or time-series models in production (forecasting, anomaly detection, trend analysis) with recalibration.
  • •Demonstrated experience evaluating, analyzing, or improving LLM/NLP systems (eval design, quality measurement, retrieval evaluation, or agent analysis).
  • •Proficiency in Python and strong SQL for querying large analytical datasets.
Experience:3+ yearsLLMsNLPTime-seriesAIOpsNOCObservabilityEnterprise IT
Education:Bachelor's
Skills:AnalysisCommunicationExperimentationProblem-solving
Tech Stack:PythonSQLLLMsNLPLLM evaluationRetrieval-augmented systemsTime-seriesAnomaly detectionForecasting

Company Brief

Science Logic
Provides an AI-driven IT operations management (ITOM) platform that delivers monitoring, event correlation, and automation to help enterprises detect, diagnose, and resolve infrastructure and application issues across hybrid and multi-cloud environments.
Industry: SaaS
Company Size: Enterprise (1,001+ employees)
Growth: Established Company
Funding: Private Equity Backed
Headquarters: Reston, United States
Founded: 2003
WebsiteLinkedIn