Data Scientist with NLP experience
.
WFP celebrates and embraces diversity. It is committed to the principle of equal employment opportunity for all its employees and encourages qualified candidates to apply irrespective of race, colour, national origin, ethnic or social background, genetic information, gender, gender identity and/or expression, sexual orientation, religion or belief, HIV status or disability.
ABOUT WFP
The United Nations World Food Programme is the world's largest humanitarian agency fighting hunger worldwide. The mission of WFP is to help the world achieve Zero Hunger in our lifetimes. Every day, WFP works worldwide to ensure that no child goes to bed hungry and that the poorest and most vulnerable, particularly women and children, can access the nutritious food they need.
ORGANIZATIONAL CONTEXT
This job is located in Rome- HQ and reports to the Head of the Use of Evaluation unit in the Office of Evaluation (OEV)
JOB PURPOSE
The World Food Programme is the world’s largest humanitarian organization, saving lives in emergencies and using food assistance to build a pathway to peace, stability and prosperity for people recovering from conflict, disasters, and the impact of climate change.
WFP’s evaluation policy (2022) includes ambitions towards promoting and enhancing the use of evaluation evidence. There is demand for increased accessibility of evaluation evidence beyond single evaluation reports, and for faster availability of the existing evidence to potential users. WFP’s corporate evaluation strategy includes provisions to share available evaluation evidence more efficiently and more broadly. Existing systems and mechanisms enabling the use of evaluation evidence beyond single evaluations involve highly manual processes including person-led review of documents and tagging to feed into databases, intended to facilitate faster extraction based on demands.
The Office of Evaluation is exploring the development of a tailored solution require the development of a pilot and engagement with identified solution providers in coordination with the Technology Division and other relevant stakeholders. The pilot will be a first phase before a preparation to a transition to production.
OEV needs the support of a data scientist with experience in Natural Learning Process (NLP) developer who is able to deliver the technical solution as per the defined requirements.
KEY ACCOUNTABILITIES (not all-inclusive)
The OEV data scientist will be responsible for the following duties
• Develop, implement, and evaluate machine learning models for NLP tasks
• Responsible for the technical design, architecture, and coding of the NLP solutions, with an immediate focus on delivering and optimizing a production-grade pilot
• Responsible for the solution evaluation, optimization, test and assisting future transition to production.
• Ensure alignment with relevant WFP standards
• Conduct research on new techniques and algorithms, and stay up-to-date with advancements in the field of NLP.
• Assist in data collection, pre-processing, and cleaning.
• Present findings to non-technical stakeholders in a clear and actionable manner.
DELIVERABLES AT THE END OF THE CONTRACT:
• Completion of a successful pilot, optimized on agreed the project’s key performance indicators, including performance and speed.
• Documentation of the lessons learnt.
• Ensure that the solution is ready to transition to full production, meeting any corporate-level standards and requirements, if/when available, to guarantee seamless integration and alignment with existing systems and processes.
• Support technical communication with the selected provider, maintaining a strong working relationship and fostering collaboration to address any issues, challenges, or opportunities that may arise during the course of the project.
STANDARD MINIMUM QUALIFICATIONS
Degree in computer science, data science, artificial intelligence, or related field
• More than 3 years of experience in NLP (must have shipped ML applications to production) and familiarity with relevant tools and libraries.
• Proficiency in Python and experience with libraries, including HuggingFace, PyTorch, or Keras.
• Demonstrable experience and detailed understanding of transformers, Large Language Models (LLM), reinforcement learning, and deep learning techniques. Direct experience with extractive search and LLM integration is a strong plus.
• Proven experience with cloud platforms, preferably GCP and Vertex AI.
• Demonstrable experience in developing production-grade apps. Experience with model monitoring and optimization, API development. Hands-on MLOPs experience would be a strong plus.
• Capacity to create simple front-end during the pilot.
• Familiarity with WFP and UN environment, experience with similar projects and understanding of evaluation are assets
• Broad and authoritative technical knowledge of Natural Language Processing
• Fully familiar with agile methodology, tools required for the development of the NLP solution (versioning, monitoring,etc.)
• Demonstrated capacity to structure projects and respect best practices, strong organizing skills and proven ability to produce results to deadlines.
• Demonstrated experience in facilitating dialogue between technical stakeholders and non-technical client, i.e. translating client into technical specifications and vice versa.
Fluency (level C) in English language. Intermediate knowledge (level B) of a second official UN language is desirable.
DEADLINE FOR APPLICATIONS
Monday 21 August
.
WFP has a zero-tolerance approach to conduct such as fraud, sexual exploitation and abuse, sexual harassment, abuse of authority and discrimination. All selected candidates will be expected to adhere to WFP’s standards of conduct and will therefore undergo rigorous background verification internally or through third parties. Selected candidates will also be required to provide additional information as part of the verification exercise. Misrepresentation of information provided during the recruitment process may lead to disqualification or termination of employment
WFP will not request payment at any stage of the recruitment process including at the offer stage. Any requests for payment should be refused and reported to local law enforcement authorities for appropriate action.
