Harika Sri

Subject Matter Expert (SME) in Statistics and AI Model Evaluation with over 7 years of experience in data analysis. Passionate about working with AI-based companies to advance Generative AI capabilities. Specializing in RLHF, SFT, and Vision-Language Models.

Proven track record of working with international clients across distributed teams, ensuring strict adherence to PST, EST, and UTC time zones. Recognized for punctuality and time management, achieving a 99% on-time delivery rate across all projects. IIT Kanpur Alumna skilled in Rubric Creation, Code Generation, and Side-by-Side (SxS) ranking.

Software Engineer Expert — Vibe Coding at Mercor (2026-02 – 2026-04)

• Designing Atomic Rubrics (8–12 tags) and calibrated prompts to train frontier models on React/HTML/JS web applications.
• Calibrating task difficulty for model pass rates: Easy 70–100%, Medium 40–70%, Hard 0–40%.
• Validating Vision-Language Models using single-shot screenshots and code inspection.

Business Analyst — AI Evaluation at Turing (2026-01 – 2026-04)

• Executed Side-by-Side (SxS) evaluations for browser-based AI agents; conducted hallucination detection using primary sources only.
• Maintained 100% guideline adherence across all evaluation tasks.

REVIEWER - STEM OCR Specialist at Mercor (2026-02 – 2026-03)

• Validating and repairing math/STEM problem-solution pairs from OCR-extracted textbooks; standardising LaTeX for AI training datasets.
• Assessing logical consistency and completeness of multi-step mathematical Chain-of-Thought (CoT) solutions.

Generalist AI Trainer at Mercor (2026-01 – 2026-02)

• Authored System Instructions, multi-turn conversations, and Golden Responses for RLHF fine-tuning of Generalist AI agents.

STEM Evaluator — Math SME at Mercor (2025-11 – 2026-01)

• Evaluated high-complexity mathematical training data for LLM optimisation; validated CoT reasoning and corrected derivation errors.

Data Annotator & Content Moderator at Innodata Inc. (2025-11 – 2025-12)

• Executed SFT annotation for ML pipelines; refined AI-generated text for fluency, bias, and PII compliance.

AI Data Evaluator — Hindi/English at OneForma (Centific) (2024-06 – 2024-08)

• Conducted RLHF Comparative Ranking (SxS) for Hindi and English AI responses; rated outputs on accuracy and cultural nuance.

REVIEWER - Prompt Engineer at Soul AI (2024-03 – 2024-05)

• Engineered and refined prompts for RLHF training of coding and reasoning models; reviewed outputs for technical accuracy.

Assistant Statistical Officer at CEGIS · Government of Telangana (2019-08 – 2025-03)

• Led evaluation of state welfare schemes using predictive modelling; architected Telangana's Digital Repository, cutting reporting time by 40%.
• Conducted comparative interstate Time-Series Analysis to support state policy decisions.

Subject Matter Expert — Statistics at Chegg Inc. (2018-05 – 2020-11)

• Authored detailed solutions for advanced Statistics and Probability at university level with high accuracy ratings.

Master of Science in Statistics – Indian Institute of Technology (IIT) Kanpur (2016 – 2018)

Bachelor of Science in Mathematics, Statistics & Computer Science – Osmania University (2014 – 2016)