Open to opportunities

Hariom Kaushal

@hariomkaushal

Message

Results-oriented Software Engineer and AI Data Specialist.

India

Message

What I'm looking for

I am looking for long-term remote roles in AI QA or data annotation, with opportunities to work in asynchronous, globally distributed teams.

As a results-oriented Software Engineer and AI Data Specialist, I bring over 3 years of remote experience in training, debugging, and evaluating large language models (LLMs). My expertise spans SWEAP-style QA tasks, including test log validation and functional testing, where I have successfully contributed to diverse AI projects for companies like Outlier, Mindrift, Remotasks, and Alignerr. My commitment to building accurate, ethical, and user-friendly AI systems drives my work.

Throughout my career, I have honed my skills in Python, Git, and various annotation tools, maintaining high throughput while consistently meeting quality thresholds. I am passionate about improving model fluency, instruction alignment, and factuality, and I actively seek remote roles in AI evaluation, model QA, or data annotation. My strong understanding of multilingual workflows and collaborative spirit make me a valuable asset to any team.

Experience

Work history, roles, and key accomplishments

Data Labeling Specialist

Alignerr

Feb 2025 - Apr 2025 (2 months)

Labeled multilingual datasets (English & Hindi) for tasks like intent recognition, entity extraction, and contextual understanding. Followed strict QA protocols and participated in peer review rounds.

Peer Review

AI Data Trainer

Outlier

Apr 2022 - Feb 2025 (2 years 10 months)

Analyzed LLM responses across tasks like summarization, reasoning, and translation for correctness, tone, and instruction alignment. Validated test logs, JSON files, and golden patches for bugs or incomplete functionality (SWEAP validation).

LLM SWEAP Validation Prompt Rewriting Quality Assurance

AI Agent Assistant

Mindrift

Nov 2024 - Jan 2025 (2 months)

Assessed real-world factual accuracy of AI outputs across domains like open-domain chat, closed QA, and dialogue summarization. Flagged hallucinations and unsupported claims, proposing prompt-level adjustments.

AI Output Assessment Hallucination Detection Prompt Engineering

Software Engineer – LLM QA

Remotasks

Dec 2023 - Dec 2024 (1 year)

Developed Python scripts to preprocess and clean LLM training datasets. Evaluated backend model outputs using logs and golden patch comparisons.

Python data preprocessing Debugging

Education

Degrees, certifications, and relevant coursework

Dr. Ram Manohar Lohia Avadh University

Bachelor of Science, Computer Science (AI & ML)

Grade: 78%

Activities and societies: Developed a bilingual chatbot capable of basic sentiment analysis and fluency evaluation. Led hands-on workshops on Python basics, data wrangling, and machine learning fundamentals.

Completed a Bachelor of Science in Computer Science with a specialization in AI & ML. Developed a bilingual chatbot capable of basic sentiment analysis and fluency evaluation as a capstone project. Led hands-on workshops on Python basics, data wrangling, and machine learning fundamentals.