Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Hello, I'm Gangula Sri Harsha, an AI Engineer and Data Scientist specializing in LLMs, prompt engineering, and model evaluation. I enjoy turning complex data into practical, scalable AI solutions using Python, PyTorch, and SQL, with a focus on refining AI outputs and improving natural language quality. I thrive in cross-functional teamwork, translating research insights into real-world products and data-driven strategies. In my recent roles, I have trained LLMs, conducted qualitative error analysis, and built end-to-end AI workflows, collaborating with product, content, and engineering teams to deliver reliable, user-friendly AI experiences.…Hello, I'm Gangula Sri Harsha, an AI Engineer and Data Scientist specializing in LLMs, prompt engineering, and model evaluation. I enjoy turning complex data into practical, scalable AI solutions using Python, PyTorch, and SQL, with a focus on refining AI outputs and improving natural language quality. I thrive in cross-functional teamwork, translating research insights into real-world products and data-driven strategies. In my recent roles, I have trained LLMs, conducted qualitative error analysis, and built end-to-end AI workflows, collaborating with product, content, and engineering teams to deliver reliable, user-friendly AI experiences.

Gangula Sri Harsha

QA Engineer, AI Prompt Engineer, Programmer, +7





Hello, I'm Gangula Sri Harsha, an AI Engineer and Data Scientist specializing in LLMs, prompt engineering, and model evaluation. I enjoy turning complex data into practical, scalable AI solutions using Python, PyTorch, and SQL, with a focus on refining AI outputs and improving natural language quality. I thrive in cross-functional teamwork, translating research insights into real-world products and data-driven strategies. In my recent roles, I have trained LLMs, conducted qualitative error analysis, and built end-to-end AI workflows, collaborating with product, content, and engineering teams to deliver reliable, user-friendly AI experiences.…Hello, I'm Gangula Sri Harsha, an AI Engineer and Data Scientist specializing in LLMs, prompt engineering, and model evaluation. I enjoy turning complex data into practical, scalable AI solutions using Python, PyTorch, and SQL, with a focus on refining AI outputs and improving natural language quality. I thrive in cross-functional teamwork, translating research insights into real-world products and data-driven strategies. In my recent roles, I have trained LLMs, conducted qualitative error analysis, and built end-to-end AI workflows, collaborating with product, content, and engineering teams to deliver reliable, user-friendly AI experiences.

Available to hire

Hello, I’m Gangula Sri Harsha, an AI Engineer and Data Scientist specializing in LLMs, prompt engineering, and model evaluation. I enjoy turning complex data into practical, scalable AI solutions using Python, PyTorch, and SQL, with a focus on refining AI outputs and improving natural language quality. I thrive in cross-functional teamwork, translating research insights into real-world products and data-driven strategies.

In my recent roles, I have trained LLMs, conducted qualitative error analysis, and built end-to-end AI workflows, collaborating with product, content, and engineering teams to deliver reliable, user-friendly AI experiences.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Beginner

Beginner

Beginner

Language

English

Fluent

Work Experience

Content Researcher (Freelance) at Asymmetric Labs

April 1, 2025 - September 1, 2025

Developed purpose-specific AI prompts for content creation, answer generation, and automated explanations, improving workflow efficiency by 25% and accelerating AI adoption across teams. Performed data analysis and visualization of 13 years of previous question papers (PYQs), extracting trends across subjects and question types using Python (Pandas, Matplotlib), enabling a 20% improvement in targeted preparation strategies. Collaborated cross-functionally with Technical and content teams to implement data-driven improvements in question frameworks and user interface, enhancing learning outcomes and engagement metrics.

Systems Engineer (Full-Time) at Tata Consultancy Services

June 1, 2020 - September 23, 2025

Executed 500+ data validation and automated test cases, improving BFSI application reliability and reducing critical defects by 30%. Collaborated with developers and analysts to streamline workflows, achieving a 20% reduction in process bottlenecks and ensuring 100% compliance with business logic. Developed 10+ domain knowledge documents and delivered presentations, enhancing stakeholder understanding and reducing system vulnerabilities.

LLM Trainer (Freelance) at OutlierAI

November 1, 2025 - Present

Conducted deep qualitative analysis of model errors and user scenarios, identifying root causes and designing implementable improvements. Produced high-quality, human-standard training data through systematic testing, evaluation, and refinement, ensuring consistency, precision, and reliability. Collaborated across product, content, and engineering teams to drive end-to-end improvements in a multi-stakeholder program context.

Content Researcher (Consultant) – Full-Time at Asymmetric Labs

April 1, 2025 - August 1, 2025

Analyzed 13 years of UPSC prelims data (1,500+ Qs) using Python to extract trends and patterns, improving curriculum alignment and preparation strategies by ~20%. Designed structured taxonomies, study guides, rubrics, and reports, translating research into practical tools and delivering clear, data-backed insights to leadership. Integrated AI-assisted workflows (prompt engineering) to streamline content creation and analysis, boosting workflow efficiency by ~25% through collaboration with CEO, content, and technical teams.

Software Engineer at Tata Consultancy Services

June 1, 2018 - June 1, 2020

Executed data validation and automated test cases, improving BFSI application reliability and reducing critical defects by 30%. Collaborated with developers and analysts to streamline workflows, achieving a 20% reduction in process bottlenecks and ensuring 100% compliance with business logic. Developed comprehensive documentation and stakeholder presentations to communicate technical requirements and implementation feasibility, ensuring alignment between business objectives and operational capacity.