Sanika T

Generated high-quality multimodal training data (text, audio, and tool-based workflows) for LLM training pipelines.

Recorded long-form (30-45 min) task demonstrations simulating real-world usage across tools such as R, VS Code, Blender, GIMP, and Apple ecosystem environments.

Applied detailed annotation protocols and QA standards to maintain clarity and reproducibility across tasks.

Evaluated LLM responses across 10+ parameters including accuracy, coherence, safety, and factual reliability.

Identified and flagged issues (hallucinations, bias, policy violations) and instruction-following gaps.

Applied rubric-based evaluation frameworks to maintain alignment across tasks.

Reviewed and calibrated peer annotations to ensure adherence to project standards.

Evaluated map-based search results for relevance, accuracy, naming, categorization, and user intent by analyzing map viewports and query context.

Provided structured feedback to support QA decisions and improve dataset quality.

Built an automated cart recovery system using Google Analytics and Zapier by analysing customer journeys, identifying key drop-off points, and translating insights into rule-based triggers and messaging flows to increase retention.

Led cross-functional initiatives across 4+ teams, expanding retail partnerships and achieving brand presence in 3 metro cities, each generating ₹3L+ in monthly sales.

Introduced automation across internal operations, reducing manual effort by 30% and improving team efficiency.

Designed a returns management system with an ongoing feedback loop, helping reduce return rates by 2% monthly and enabling a faster return/exchange system to address customer pain-points.

Completed graduate coursework for a Master of Statistics and Data Science focusing on data management, probability & statistics, linear models, R programming, and Bayesian inference.

Activities and societies: Certificate Programme in Data Science: Basic & Advanced Excel, VBA, R Fundamentals, Statistical Methods

Postgraduate Diploma in Data Science covering data mining, big data analytics, predictive modelling, time series, statistical inference, exploratory data analysis, and programming in R and Python.

Selected from a pool of 66,000+ candidates worldwide for Stanford’s Code In Place (CS106A), an intensive, 6-week
program covering core programming concepts and problem-solving in Python.

Activities and societies: INSPIRE Scholarship Programme recipient

Bachelor of Science in Statistics with coursework in probability theory, calculus, linear algebra, statistical estimation, operations research, ANOVA, and programming in Java and C++.