For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

Invite to Job
Sarha Pompili

Sarha Pompili

AI trainer - Project management, workflow and automation

FRANCE flag
Paris, France
$55.00/hrEntry LevelOther

Key Skills

Software

Other

Top Subject Matter

Project management
Workflow and process engineering
Generative AI and Automation

Top Data Types

DocumentDocument
TextText
AudioAudio

Top Task Types

Prompt Response Writing SFT

Freelancer Overview

My experience in AI training data stems from a combination of professional project delivery and deliberate hands-on practice. As Project Director at Brink's, I designed and delivered an AI-powered invoicing automation workflow, which required auditing data quality, structuring input/output formats, and building validation logic to ensure the model worked on clean, reliable data. More recently, I have independently built four AI workflow projects across different platforms (Alteryx, n8n, Call Rounded, Relevance.AI, Make), including a generative AI chatbot, a voice agent, and a multi-agent content system. Each project required designing realistic scenarios, anticipating edge cases, and evaluating whether the AI output was operationally viable, not just technically coherent. What sets my profile apart is the combination of 10+ years of real-world process expertise with active AI practice. I understand how processes actually work in companies, where data breaks, where human judgment is irreplaceable, and where AI creates genuine value versus a false sense of control. This grounding allows me to assess AI-generated content not just for technical accuracy, but for business relevance and operational feasibility. I have also trained myself to evaluate AI outputs comparatively, for instance, building the same use case on two different platforms to understand their trade-offs. This is the kind of critical, calibrated judgment that AI training data quality requires.

Entry LevelFrenchEnglishSpanish

Labeling Experience

Prompt Engineering, AI Chatbot and Multi-Agent Content System Practice Projects

OtherTextPrompt Response Writing SFT
Built several independent generative AI and conversational AI projects involving writing, refining, and evaluating AI prompts and responses. Developed a generative AI chatbot for a travel agency, a voice agent for restaurant order-taking, and multi-agent social media content systems. Focused on prompt engineering, multi-tool orchestration, AI workflow automation, and content adaptation for platform consistency. • Wrote and iteratively improved prompt instructions. • Evaluated and adjusted AI-generated text outputs. • Coordinated agent responses for consistent brand tone. • Built workflows to integrate AI responses across use cases.

Built several independent generative AI and conversational AI projects involving writing, refining, and evaluating AI prompts and responses. Developed a generative AI chatbot for a travel agency, a voice agent for restaurant order-taking, and multi-agent social media content systems. Focused on prompt engineering, multi-tool orchestration, AI workflow automation, and content adaptation for platform consistency. • Wrote and iteratively improved prompt instructions. • Evaluated and adjusted AI-generated text outputs. • Coordinated agent responses for consistent brand tone. • Built workflows to integrate AI responses across use cases.

2025 - 2026

AI Output Evaluation & Data Quality Management – Invoicing Automation

DocumentData Collection
Designed and implemented an AI-powered workflow for automating invoicing tasks and evaluating AI-generated outputs. Managed data validation, developed exception handling logic, and ran automated data quality checks to identify and manage anomalies. Evaluated and improved the quality and business relevance of AI-generated invoice-ready files for a logistics company. • Built processes to review and rate output accuracy. • Developed workflows for ongoing data quality improvement. • Used structured evaluation criteria for invoice automation. • Enhanced input data detection and error correction.

Designed and implemented an AI-powered workflow for automating invoicing tasks and evaluating AI-generated outputs. Managed data validation, developed exception handling logic, and ran automated data quality checks to identify and manage anomalies. Evaluated and improved the quality and business relevance of AI-generated invoice-ready files for a logistics company. • Built processes to review and rate output accuracy. • Developed workflows for ongoing data quality improvement. • Used structured evaluation criteria for invoice automation. • Enhanced input data detection and error correction.

2021 - 2022

Education

E

EM Lyon Business School

Master of Science, Entrepreneurship and Innovation Management

Master of Science
2013 - 2014
U

Universite Paris-Dauphine

Master of Science, Marketing and Strategy

Master of Science
2012 - 2013

Work History

B

Brink's Indonesie

Head of Implementation Proejcts

Jakarta
2025 - Present
B

Brink's

Project Director

Paris
2022 - 2025