Data Expert - Medical/Public Health

Wadhwani AI; Delhi; On-site; INR 360,000 - 600,000; Full time

Job summary

Wadhwani AI is a non-profit institute building applied AI solutions to address public health, agriculture, education, and urban development in underserved communities. The team collaborates with governments, social sector organizations, and research institutions to deploy practical AI solutions for measurable impact.

The role focuses on programmatic evaluation of an AI assistant, QA pair validation, and converting documents to structured Markdown while preserving content and structure.

Qualifications

• 1-2 years of experience in public health preferred.
• Basic understanding of digital and AI tools, LLMs and generative AI chatbots.
• Excellent written and verbal communication skills.
• Clear and concise visual and written documentation skills for communicating system enhancements.

Responsibilities

• Create comprehensive evaluation datasets covering variations in input text and domain knowledge.
• Perform programmatic evaluation of the AI assistant's responses and provide feedback with reasoning.
• Validate accuracy of QA pairs against documents/pages and modify or create QA pairs to ensure full coverage.

Skills

Public health exposure

Digital & AI basics

Communication

Documentation

Marathi or Odia language

Tools

Excel

Google Sheets

Job description

Wadhwani AI is a non-profit institute building and deploying applied AI solutions to solve critical issues in public health, agriculture, education, and urban development in underserved communities in the global south. We collaborate with governments, social sector organizations, academic and research institutions, and domain experts to identify real-world problems, and develop practical AI solutions to tackle these issues with the aim of making a substantial positive impact.

We have over 30+ AI projects supported by leading philanthropies such as Gates Foundation, Google.org and Walmart. With a team of over 200 professionals, our expertise encompasses AI/ML research and innovation, software engineering, domain knowledge, design and user research.

Job Description

Programmatic Support for Evaluating AI Assistant

• Create comprehensive, diverse evaluation datasets that incorporate variations in input text and cover both field-level and domain-level knowledge.
• Perform programmatic evaluation of the AI assistant's responses.
• Provide detailed feedback on model-generated answers using reference data, and offer clear reasoning for the expected answers.

Validation of Question-Answer Pairs for the AI Assistant

• Verify the accuracy of information within QA pairs against specified documents and pages. Validate QA pairs for correctness, completeness, and relevance based on document content.
• Modify existing QA pairs or create new ones to ensure full coverage of the document text. Highlight relevant passages within documents to ensure QA-pair coverage.

Document Digitization Using Markdown Syntax

• Conduct a document reconnaissance to understand nuances and structure. Convert documents to Markdown syntax while preserving the original structure and content.
• Reorganize and rewrite information in a natural, logical sequence to ensure proper flow (top to bottom, left to right, or as specified).
• Maintain consistent formatting, including correct headings, sections, and page numbers. Convert text into images, flowcharts, and tables to markdown syntax and place it accurately within the corresponding sections and headings.
  

Requirements

• 1-2 years of experience in public health is preferred, along with familiarity with Frontline Workers(ASHAs, AWWs, ANMs, Supervisors, etc) roles and the HealthICDS/WCD ecosystem.
• Have in depth understanding of various guidlines, documents developed for ASHA/AWW/FLWs by MoHFW, MoWCD and line departments
• Basic understanding of digital and AI tools, LLMs and generative AI Chatbots
• Experience working with or training frontline healthcare workers is an asset.
• Strong attention to detail to ensure accuracy and completeness of QA pairs and document digitisation.
• Comfort with multi-tasking in a split screen setup, using Excel or Google Sheets, and performing multiple copy-paste operations.
• Excellent written and verbal communication skills to document updates and provide clear feedback.
• Clear and concise visual and written documentation skills to effectively communicate system enhancements to the development team.
• Basic knowledge and a keen interest to learn more about AI deployments for social impact.
• Proficiency in both spoken and written Marathi or Odia is required.