Lakita Vashisth

Key Skills

Software

Data Annotation Tech

iMerit

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Top Task Types

Classification

Computer Programming Coding

Evaluation Rating

Prompt Response Writing SFT

Question Answering

Text Summarization

Freelancer Overview

I am an experienced AI prompt engineer and mathematics content creator with a strong background in designing, evaluating, and optimizing training data for large language models. My expertise includes developing high-quality prompts, question banks, and evaluation frameworks for mathematics domains such as algebra, calculus, probability, and statistics, using advanced techniques like chain-of-thought, Socratic questioning, and few-shot/zero-shot prompting. I have hands-on experience collaborating with educators and AI researchers to ensure curriculum alignment and accuracy, as well as evaluating model outputs for safety and correctness. Skilled in tools like OpenAI GPT, Anthropic Claude, and Gemini, I have contributed to multilingual and multicultural prompt evaluation projects, as well as the creation of prompt libraries for reasoning, summarization, and data extraction. My academic background in applied mathematics and computer science enables me to simplify complex concepts for both learners and AI systems, ensuring high-quality, interpretable AI training data.

Labeling Experience

Prompt Framework Developer (Soul AI)

Data Annotation TechTextPrompt Response Writing SFT

At Soul AI, I developed and optimized structured prompt frameworks for AI tasks, focusing on reasoning, summarization, and data extraction. I applied zero-shot and few-shot prompt engineering to enhance model generalization and decision-making accuracy. My work involved testing and refining prompt effectiveness within AI reasoning workflows. • Created, tested, and iterated on prompt sets for various LLM-based tasks • Enhanced model capabilities in logical reasoning and text summarization • Applied few-shot learning methods to new problem domains • Delivered feedback on model performance to improve prompt structures.

At Soul AI, I developed and optimized structured prompt frameworks for AI tasks, focusing on reasoning, summarization, and data extraction. I applied zero-shot and few-shot prompt engineering to enhance model generalization and decision-making accuracy. My work involved testing and refining prompt effectiveness within AI reasoning workflows. • Created, tested, and iterated on prompt sets for various LLM-based tasks • Enhanced model capabilities in logical reasoning and text summarization • Applied few-shot learning methods to new problem domains • Delivered feedback on model performance to improve prompt structures.

2023

Prompt Evaluator (Telus International AI)

TelusTextEvaluation Rating

For Telus International AI, I engineered, tested, and evaluated multilingual prompts to ensure their accuracy, cultural appropriateness, and sensitivity in AI systems. I reviewed AI-generated outputs across languages, conducting detailed alignment checks with user intent and task requirements. My work contributed directly to AI model quality and inclusivity. • Developed multilingual prompts for diverse cultural contexts • Performed systematic prompt evaluation and output ratings • Ensured content correctness and alignment with task guidelines • Collaborated with stakeholders for continuous improvement in evaluation methods.

For Telus International AI, I engineered, tested, and evaluated multilingual prompts to ensure their accuracy, cultural appropriateness, and sensitivity in AI systems. I reviewed AI-generated outputs across languages, conducting detailed alignment checks with user intent and task requirements. My work contributed directly to AI model quality and inclusivity. • Developed multilingual prompts for diverse cultural contexts • Performed systematic prompt evaluation and output ratings • Ensured content correctness and alignment with task guidelines • Collaborated with stakeholders for continuous improvement in evaluation methods.

2023

Prompt Engineer (Outlier)

Data Annotation TechTextPrompt Response Writing SFT

As a Prompt Engineer at Outlier, I built and optimized prompts for mathematics, including algebra, calculus, probability, and statistics across multiple linguistic settings. I designed structured prompt libraries for tasks like mathematical reasoning, dialogue, summarization, and multilingual evaluation, focusing on clarity, safety, and accuracy. My responsibilities included developing evaluation frameworks, implementing few-shot and chain-of-thought strategies, and reviewing AI model outputs for correctness and alignment. • Developed and evaluated advanced prompt engineering tasks supporting LLM data training • Implemented prompt techniques for AI interpretability and multi-step reasoning • Designed prompt sets specifically for summarization, data extraction, and reasoning • Collaborated with teams to ensure curriculum alignment and best practices documentation.

As a Prompt Engineer at Outlier, I built and optimized prompts for mathematics, including algebra, calculus, probability, and statistics across multiple linguistic settings. I designed structured prompt libraries for tasks like mathematical reasoning, dialogue, summarization, and multilingual evaluation, focusing on clarity, safety, and accuracy. My responsibilities included developing evaluation frameworks, implementing few-shot and chain-of-thought strategies, and reviewing AI model outputs for correctness and alignment. • Developed and evaluated advanced prompt engineering tasks supporting LLM data training • Implemented prompt techniques for AI interpretability and multi-step reasoning • Designed prompt sets specifically for summarization, data extraction, and reasoning • Collaborated with teams to ensure curriculum alignment and best practices documentation.

2022

PROMPT ENGINEER

Data Annotation TechTextQuestion AnsweringText Summarization

I have strong experience in AI training data and data labeling, gained through large-scale projects with Outlier,Telus International AI, and Soul AI. My work has spanned building structured prompt libraries, multilingual evaluation datasets, and nreasoning/summarization frameworks to train and fine-tune AI models. I performed data labeling tasks such as annotating and validating AI outputs for accuracy, logical consistency, safety, and cultural sensitivity, along with tagging prompts and responses for structured datasets.These projects supported thousands of labeled samples across mathematics, dialogue, and moderation domains, executed in collaboration with educators and AI researchers. To maintain quality, I adhered to multi-level review cycles, inter-annotator agreement checks, and benchmark-based validations, ensuring reliability and bias reduction. With expertise in applied mathematics and prompt engineering, I bring both technical depth and precision to AI training datadevelopment

I have strong experience in AI training data and data labeling, gained through large-scale projects with Outlier,Telus International AI, and Soul AI. My work has spanned building structured prompt libraries, multilingual evaluation datasets, and nreasoning/summarization frameworks to train and fine-tune AI models. I performed data labeling tasks such as annotating and validating AI outputs for accuracy, logical consistency, safety, and cultural sensitivity, along with tagging prompts and responses for structured datasets.These projects supported thousands of labeled samples across mathematics, dialogue, and moderation domains, executed in collaboration with educators and AI researchers. To maintain quality, I adhered to multi-level review cycles, inter-annotator agreement checks, and benchmark-based validations, ensuring reliability and bias reduction. With expertise in applied mathematics and prompt engineering, I bring both technical depth and precision to AI training datadevelopment

2022 - 2025

Education

Work History