Hardik Kalra

Team Lead - AI Training at Turing - OpenAI US - Remote

Led and managed 4 pod leads (25+ trainers) on OpenAI’s robotics and gaming video annotation projects, delivering large-scale multimodal datasets with atomic subtasks, intent captions, rationales, and structured summaries for model training. Designed annotation guidelines for robotics and gaming action sequences; implemented multi-layer audits to boost annotation accuracy and consistency by 25%.

Pod Lead - AI Training at Turing - Amazon US - Remote

Managed a pod of 8 AI trainers for Amazon’s proprietary LLMs on graphs, scientific visuals, and analytical plots; created complex reasoning tasks; oversaw pod-level evaluations and quality checks, improving accuracy and efficiency across multimodal datasets.

AI Developer - LLM Trainer at Turing - Google US - Remote

Fine-tuned LLMs using supervised fine-tuning on advanced mathematics problems from JEE and Olympiad datasets; designed human-readable solutions and led AI data training workflows; labeled and audited datasets using CVAT and other annotation tools.

Data Scientist at Medlr

Developed the RapidFuzz technology algorithm using NLP to reduce medication errors by 30% and processed data from 697,620 medicines across top pharmacies; automated web scraping from Tata1mg, Pharmeasy, and Netmeds and integrated into ElasticSearch for insights; designed a scheduling system for periodic data updates and expanded dataset by 30% year-on-year.

Prompt Engineer - AI Trainer at Outlier AI US - Remote

Designed, tested, and refined prompts for AI models across multiple projects, ensuring high accuracy in STEM, coding, general science, and knowledge-based problem solving; developed optimized prompt structures to improve reasoning, clarity, and contextual relevance; evaluated models for Indian English and conducted RLHF/SFT analyses to scale prompt-based training workflows.

Prompt Engineer - AI Tutor at Soul AI US - Remote

Developed and fine-tuned interactive prompts for LLMs; collaborated with SMEs to integrate industry-specific terminology; conducted data annotation and labeling for NLP tasks; monitored prompt performance, analyzed feedback, and iteratively optimized AI responses.

Team Lead - AI Training at Turing

Led and managed 4 pod leads (25+ trainers) on OpenAI’s robotics and gaming video annotation projects, delivering large-scale multimodal datasets with atomic subtasks, intent captions, rationales, and structured summaries for model training. Designed annotation guidelines for robotics and gaming action sequences, and implemented multi-layer audits to boost annotation quality by 25%. Reported cross-pod performance and milestones to senior management and aligned outputs with downstream model goals.

Pod Lead - AI Training at Turing

Managed a pod of 8 AI trainers for Amazon’s proprietary LLMs on graphs, scientific visuals, and analytical plots; created complex reasoning tasks, oversaw pod-level evaluations, and applied error analysis to improve accuracy in visual reasoning tasks. Mentored trainers and ensured quality checks across multimodal datasets.

AI Developer - LLM Trainer at Turing

Fine-tuned LLMs using Supervised Fine-Tuning (SFT) on advanced mathematics problems from JEE and Olympiad-level datasets; designed clear human-readable solutions by combining prompt engineering with mathematical research; led AI data training workflows and audited outputs for consistency and pedagogy.

Data Scientist at Medlr

Developed the RapidFuzz technology algorithm using NLP, reducing medication errors by 30% and processing data from 697,620 medicines across top pharmacies; automated web scraping from Tata1mg, Pharmeasy, and Netmeds; integrated data into ElasticSearch and designed a scheduling system for periodic updates; utilized multithreading to boost throughput.

Prompt Engineer - AI Trainer at Outlier AI

Designed, tested, and refined prompts for AI models across Kepler-RLHF, Psychic – Kadabra, Cypher-RLHF, and Bee-SFT-Multilingual projects; developed optimized prompt structures to enhance reasoning, clarity, and contextual relevance; evaluated Indian English (IN) and refined linguistic precision for inclusive interactions.

AI Tutor - Prompt Engineer at Soul AI

Developed and fine-tuned interactive prompts for large language models; collaborated with subject matter experts to incorporate industry-specific terminology; conducted data annotation and labeling for NLP tasks and image datasets; monitored prompt performance and iterated to improve AI responses.

Team Lead - AI Training at Turing - OpenAI US

Led and managed 4 pod leads (25+ trainers) on OpenAI’s robotics and gaming video annotation projects; delivered large-scale multimodal datasets with atomic subtasks, intent captions, rationales, and structured summaries for model training. Designed annotation guidelines for robotics and gaming action sequences; implemented multi-layer audits that boosted annotation accuracy and consistency by 25%. Mentored onboarding and reported performance metrics to senior management; partnered with research teams to refine edge-case outputs for downstream model development.

Pod Lead - AI Training at Turing - Amazon US

Managed a pod of 8 AI trainers, overseeing end-to-end annotation workflows for Amazon’s proprietary LLMs on graphs, scientific visuals, and analytical plots. Created complex reasoning tasks to enhance model capabilities in visual reasoning and data analysis. Led pod-level evaluations, refined trainer prompts, and applied error analysis to systematically improve accuracy and efficiency; collaborated with Amazon’s research team to align dataset outputs with evolving project objectives.

AI Developer - LLM Trainer at Turing - Google US

Fine-tuned LLMs using Supervised Fine-Tuning (SFT) on advanced mathematics problems from JEE and Olympiad-level datasets; designed clear, human-readable solutions by combining prompt engineering with rigorous mathematical research; led AI data training workflows, creating, labeling, and auditing high-quality datasets using CVAT to enhance model comprehension of mathematical expressions and logic. Reviewed outputs from junior engineers to ensure consistency and pedagogical alignment.

Data Scientist at Medlr

Developed the RapidFuzz technology algorithm using NLP, reducing medication errors by 30% and processing data from 697,620 medicines across top pharmacies. Automated web scraping for pharmaceutical data from Tata1mg, Pharmeasy, and Netmeds, integrating it into Elasticsearch to enhance data-driven insights. Designed a scheduling system for periodic data updates, expanding the dataset by 30% year-on-year and tracking price changes. Leveraged multithreading to accelerate data extraction and throughput.

Prompt Engineer - AI Trainer at Outlier AI

Designed, tested, and refined prompts for AI models across projects, including Kepler-RLHF, Psychic – Kadabra, Cypher-RLHF, and Bee-SFT-Multilingual; developed optimized prompt structures to enhance AI-generated responses, improving logical reasoning, clarity, and contextual relevance across STEM, coding, general science, and knowledge-based tasks. Evaluated AI models for Indian English, applying RLHF and SFT to improve model performance. Collaborated with cross-functional teams to scale prompt-based training workflows.

Prompt Engineer - AI Tutor at Soul AI

Developed and fine-tuned interactive prompts for large language models (LLMs), improving response accuracy and relevance by 25%; collaborated with subject matter experts to integrate industry-specific terminology, boosting prompt effectiveness and user engagement. Conducted data annotation and labeling for text-based NLP tasks and image datasets; monitored prompt performance in real-world scenarios and iteratively optimized AI responses.

Team Lead - AI Training at Turing (OpenAI) - Remote

Led and managed 4 pod leads (25+ trainers) on robotics and gaming video annotation projects, delivering large-scale multimodal datasets with atomic subtasks, intent captions, rationales, and structured summaries for model training. Designed annotation guidelines for robotics and gaming action sequences to translate movement data into model-ready formats for downstream multimodal reasoning. Benchmarked dataset quality against baselines and best practices, introducing multi-layer audits and feedback loops that boosted annotation accuracy and consistency by 25%. Mentored onboarding and workflow optimization; reported cross-pod performance to senior management. Collaborated with research teams to refine templates and edge cases for alignment with model development.

Pod Lead - AI Training at Turing - Amazon

Managed a pod of 8 AI trainers for Amazon’s proprietary LLMs on graphs, scientific visuals, and analytical plots. Created complex reasoning tasks involving visual approximation, multi-step analysis, trend detection, and value inference to improve model analytical problem solving. Oversaw pod-level evaluations and error analysis to enhance accuracy in visual comprehension. Collaborated with Amazon research to align outputs with project objectives. Mentored trainers through onboarding and quality checks to improve accuracy and efficiency.

AI Developer - LLM Trainer at Turing - Google

Fine-tuned LLMs using supervised fine-tuning on advanced mathematics problems from JEE and Olympiad-level datasets, optimizing structured reasoning and final-answer accuracy. Designed clear human-readable solutions by combining prompt engineering with mathematical research, leveraging GPT-4o and O-1 Preview. Led AI data training workflows; created, labelled, and audited high-quality datasets using CVAT and other annotation tools to enhance model comprehension of mathematical expressions and logic. Reviewed outputs from junior engineers for consistency and pedagogical alignment.

Data Scientist at Medlr

Developed RapidFuzz technology algorithm using NLP, reducing medication errors by 30% and processing data from 697,620 medicines across top pharmacies for precision and efficiency. Automated web scraping for pharmaceutical data from Tata1mg, Pharmeasy, Netmeds; integrated into ElasticSearch to enhance data-driven insights. Designed a scheduling system for periodic data updates, expanding the dataset by 30% YoY; leveraged multithreading to accelerate data extraction.

Prompt Engineer - AI Trainer at Outlier AI

Designed, tested, and refined prompts for AI models across multiple projects (Kepler-RLHF, Psychic – Kadabra, Cypher-RLHF, Bee-SFT-Multilingual), ensuring high accuracy in STEM, coding, and knowledge-based problem-solving. Developed optimized prompt structures to enhance AI-generated responses and evaluated models for Indian English linguistic precision and cultural context. Conducted RLHF, SFT, reward modelling; scaled prompt-based training workflows in collaboration with cross-functional teams.

Prompt Engineer - AI Tutor at Soul AI

Developed and fine-tuned interactive prompts for LLMs, improving response accuracy and relevance by 25%. Collaborated with subject-matter experts to integrate industry terminology; conducted data annotation for NLP tasks and image datasets; monitored prompt performance and iteratively optimized AI-generated responses, streamlining data collection and decision-making.

Team Lead - AI Training at Turing - OpenAI US

Led and managed 4 pod leads (25+ trainers) on OpenAI’s robotics and gaming video annotation projects, delivering large-scale multimodal datasets with atomic subtasks, intent captions, rationales, and structured summaries for model training. Designed and enforced annotation guidelines for robotics and gaming action sequences; benchmarked dataset quality; introduced multi-layer audits and targeted feedback loops that boosted annotation accuracy and consistency by 25%. Mentored pod leads through onboarding, workflow optimization, and advanced annotation techniques, while reporting cross-pod performance metrics, risks, and delivery milestones directly to senior managers. Partnered with OpenAI’s research teams to refine annotation templates and integrate robotics and gaming-specific edge cases, aligning data outputs with downstream model development and multimodal reasoning goals.

Pod Lead - AI Training at Turing - Amazon US

Managed a pod of 8 AI trainers, reporting to the Team Lead, and oversaw end-to-end annotation workflows for Amazon’s proprietary LLMs on graphs, scientific visuals, and analytical plots. Created complex reasoning tasks involving visual approximation, multi-step analysis, trend detection, and value inference—enhancing the model’s ability to solve exam-style and real-world analytical problems. Oversaw pod-level evaluations of model performance, refining trainer prompts and applying error analysis to systematically improve accuracy in visual comprehension and reasoning tasks. Collaborated with Amazon’s research team to relay pod-level progress, integrate multimodal feedback, and align dataset outputs with evolving project objectives. Mentored and supported pod trainers through onboarding, workflow optimization, and detailed quality checks—improving annotation accuracy, consistency, and efficiency across multimodal datasets.

AI Developer - LLM Trainer at Turing - Google US

Fine-tuned LLMs using Supervised Fine-Tuning (SFT) on advanced mathematics problems from JEE and Olympiad-level datasets, optimizing structured reasoning and final-answer accuracy. Designed clear, human-readable solutions by combining advanced prompt engineering with deep mathematical research, leveraging models such as GPT-4o and O-1 Preview to improve clarity, precision, and step-by-step logic. Led AI data training work flows—created, labelled, and audited high-quality datasets using CVAT and other annotation tools to enhance model comprehension of mathematical expressions and logic. Reviewed and refined outputs from junior engineers and new joiners, ensuring consistency, correctness, and alignment with pedagogical and project standards.

Data Scientist at Medlr

Developed the RapidFuzz technology algorithm using NLP, reducing medication errors by 30% and processing data from 697,620 medicines across top pharmacies for precision and efficiency. Automated web scraping for pharmaceutical data from platforms like Tata1mg, Pharmeasy, and Netmeds, integrating it into ElasticSearch to enhance data-driven insights. Designed a scheduling system for periodic data updates, expanding the dataset by 30% year-on-year while tracking price changes. Leveraged multithreading to accelerate data extraction and improve throughput, driving greater efficiency and productivity.

Prompt Engineer - AI Trainer at Outlier AI

Designed, tested, and refined prompts for AI models across projects, including Kepler-RLHF, Psychic – Kadabra, Cypher-RLHF, and Bee-SFT-Multilingual, ensuring high accuracy in STEM, coding, general science, and knowledge-based problem-solving. Developed optimized prompt structures to enhance AI-generated responses, improving logical reasoning, clarity, and contextual relevance across diverse academic and practical domains. Evaluated AI models for Indian English (IN), refining linguistic precision and cultural context to support inclusive and accurate user interactions. Conducted systematic performance analyses using RLHF, SFT, reward modelling, and advanced prompting techniques to improve model consistency, response quality, and user alignment. Collaborated with cross-functional teams to scale and enhance prompt-based training workflows, contributing to robust advancements in AI-driven cognitive capabilities.