Anshika P.

Key Skills

Software

Don't disclose

Top Subject Matter

LLM output evaluation and quality assurance

Prompt engineering evaluation for LLMs

AI text quality labeling and review

Top Data Types

Text

Top Task Types

RLHF

Function Calling

Freelancer Overview

AI Chatbot Workflow Testing & Evaluation (Independent Research Project). Brings 4+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose, Make, and Zapier. Education includes Bachelor of Technology, ITM University Gwalior (2027). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and RLHF.

Labeling Experience

AI Chatbot Workflow Testing & Evaluation (Independent Research Project)

Don't discloseText

Evaluated AI chatbot outputs across multiple LLMs by designing structured test cases and assessing conversational quality. Documented recurring failure patterns such as hallucinations, instruction drift, and context loss with severity ratings. Applied a repeatable multi-criterion rubric to consistently score prompt-response pairs at scale. • Built 40+ structured test cases for coherence, factual accuracy, and task-completion quality • Produced actionable evaluation reports with severity ratings for model-specific failures • Used a 5-criterion rubric (accuracy, relevance, fluency, safety, usefulness) across 100+ evaluations • Benchmarked ChatGPT (GPT-4), Claude 3, and Gemini 1.5 Pro

2024 - Present

No-Code AI Automation Workflows (Exploratory Project)

TextFunction Calling

Built and tested no-code AI automation workflows that generate and refine text outputs for real use cases. Iteratively evaluated workflow reliability and output quality by adjusting AI instructions to reduce off-topic responses. Documented design decisions, edge cases, and optimization steps to improve consistency of generated content. • Created 5+ no-code automation workflows using Make, Zapier, and AI-integrated tools • Refined prompts/instructions to improve reliability and consistency of outputs • Reduced off-topic responses by >40% through iterative testing • Documented workflow design decisions, edge cases, and optimization steps

2023 - 2023

AI Content Quality Assessment Practice (Self-Directed Learning)

Don't discloseText

Practiced AI content quality assessment by reviewing and rating large volumes of AI-generated text across multiple domains. Flagged ambiguous, misleading, or low-quality outputs and provided written justifications following standardized quality dimensions. Maintained detailed annotation logs to demonstrate consistent judgment under clearly defined guidelines. • Rated 200+ AI-generated text samples (creative writing, factual Q&A, and code explanation) • Wrote justifications for flagged outputs to simulate reviewer labeling workflows • Followed standardized quality dimensions for consistent evaluation • Maintained annotation logs supporting high-volume, guideline-based assessment

2023 - 2023

Prompt Comparison & Optimization Study (Personal Initiative)

Don't discloseTextRLHF

Performed prompt-comparison testing by running systematic A/B analyses of zero-shot versus few-shot strategies on major LLMs. Logged performance differences in a structured tracker to support repeatable evaluation and analysis. Summarized findings to identify prompt patterns that improved response specificity. • Conducted A/B analysis of zero-shot vs. few-shot prompts across 3 LLMs • Logged results in a structured spreadsheet tracker for comparison • Assessed qualitative performance differences using consistent scoring • Produced a shareable 12-page evaluation report of observed improvements

2023 - 2023

Education

Work History