Background

The Alliance of Bioversity International (www.bioversityinternational.org) and the International Center for Tropical Agriculture (CIAT) (www.ciat.cgiar.org) delivers research-based solutions that harness agricultural biodiversity and sustainably transform food systems to improve people’s lives. Alliance solutions address the global crises of malnutrition, climate change, biodiversity loss, and environmental degradation.

With novel partnerships, the Alliance generates evidence and mainstreams innovations to transform food systems and landscapes so that they sustain the planet, drive prosperity, and nourish people in a climate crisis.

The Alliance is part of CGIAR, a global research partnership for a food-secure future.

The NDIZI (NLP to Develop and Innovate Zero-shot Intelligence) project seeks a dedicated Machine Learning Operations (MLOps) Consultant (“the consultant”) to support the development, deployment, and operationalization of machine learning systems powering SIKIA, a voice-first and multimodal AI platform for conversational data collection and analysis, and related AI-driven research workflows. The consultant will work across speech, NLP, multimodal AI, and agentic pipelines, helping transition models and data systems from research prototypes into reliable, scalable solutions suitable for real-world field deployment.

We are building these systems to enable crop improvement systems to bridge formal breeding processes happening within controlled experimental trials with real-world environments on-farm. Principal use-cases include identifying farmer preferences, assessing plant disease occurrence and scoring, and environmental response - particularly related to climate adaptation. Outputs from this consultancy will be used to address these use-cases.

About the consultancy:

The consultant will support end-to-end MLOps workflows spanning data ingestion, validation, dataset versioning, model training, evaluation, deployment, monitoring, and continuous improvement across both cloud and edge environments. The role will involve close collaboration with research machine learning, software engineering, product, and field teams to ensure systems are robust, maintainable, and aligned with project needs.

The consultant will also support integration of ML systems within the SIKIA platform, including deployment workflows connecting mobile applications, cloud infrastructure, speech and multimodal pipelines, disease detection and severity scoring workflows, backend services, and FAIRGrounds-integrated systems. In addition, the role will contribute to strengthening best practices around experiment tracking, model governance, CI/CD workflows, deployment automation, and ML system monitoring across NDIZI infrastructure.

This is a 11-month full-time consultancy position based either in Arusha, Tanzania or Nairobi Kenya.

Requirements

Speech and NLP Systems Refinement

• Support deployment and optimization of multilingual ASR pipelines for cloud and mobile environments.
• Develop workflows for speech data ingestion, transcription, evaluation, and continuous model improvement.
• Implement automated retraining and fine-tuning pipelines using newly collected field data.
• Support deployment of LLM-based workflows for conversational analysis and trait extraction.
• Monitor model performance, latency, reliability, and drift under field conditions.
• Optimize inference workflows for low-connectivity and resource-constrained environments.

Multimodal Pipeline Development and Deployment

• Support development of multimodal pipelines linking speech, transcripts, metadata, and field images collected through SIKIA.
• Implement workflows for multimodal data validation, synchronization, storage, annotation, and dataset versioning.
• Support training, deployment, and evaluation of multimodal and visual-language AI models.
• Develop scalable workflows for managing multimodal datasets and model outputs across cloud infrastructure.
• Support benchmarking, reproducibility, and optimization of multimodal AI pipelines for field deployment.

Disease Detection and Severity Scoring

• Support development and deployment of AI workflows for disease detection and severity scoring using field images and multimodal data.
• Implement data and evaluation pipelines for disease annotation, validation, benchmarking, and continuous model improvement.
• Support integration of disease scoring workflows within the SIKIA and ONA platforms for field-based data collection and analysis.

MLOps Infrastructure and SIKIA Integration

• Develop CI/CD pipelines for model training, evaluation, and deployment.
• Manage experiment tracking, model registries, and dataset versioning workflows.
• Implement monitoring and logging across ML services.
• Collaborate with the software development team to integrate RAG pipelines into an agentic deployment structure.
• Support deployment of ML services on GCP, FAIRGrounds, and related infrastructure.
• Support integration of ML services within the SIKIA platform across mobile, backend, API, and cloud systems.
• Ensure compliance with data governance, security, and responsible AI requirements.

Deliverables and Payment Schedule:

Deliverable 1:Inception Report and Technical Workplan (Month 1, ~Week 4)

Develop an inception report outlining the technical approach, deployment priorities, infrastructure requirements, integration roadmap, and detailed 11-month workplan for MLOps, speech, multimodal, and disease-scoring workflows across the SIKIA platform.

Deliverable 2: Model Development, Evaluation, and Infrastructure Setup (Month 5, ~Week 20)

Establish core MLOps infrastructure and workflows, including CI/CD pipelines, experiment tracking, dataset versioning, model registries, and monitoring systems. Support development, evaluation, and optimization of speech, multimodal, and disease-scoring models across cloud and edge environments.

Deliverable 3: SIKIA AI Pipeline Integration and Deployment (Month 8, ~Week 32)

Deliver integrated deployment workflows connecting ML services with SIKIA mobile applications, APIs, backend systems, and FAIRGrounds-integrated infrastructure. Provide operational pipelines for speech processing, conversational analysis, multimodal workflows, and disease-scoring services, including deployment documentation and integration support.

Deliverable 4: Monitoring, Evaluation, and Optimization Framework (Month 10, ~Week 40)

Implement monitoring, logging, benchmarking, and evaluation workflows for deployed ML systems, including ASR performance tracking, multimodal pipeline evaluation, disease-scoring validation, and model drift monitoring. Provide recommendations for optimization, scalability, and field deployment improvements.

Deliverable 5: Final Technical Report and Handover Package (Month 11, end of assignment)

Submit a final technical report summarizing completed workflows, deployed infrastructure, system performance, key learnings, and recommendations for future scaling and maintenance. Deliver finalized documentation, deployment guides, pipeline configurations, and knowledge-transfer materials for internal teams.

Education:

Master’s degree in computer science, Data Science, Artificial Intelligence, Software Engineering, or a related field.

Experience:

At least 3 years of experience in Machine Learning engineering, MLOps, or deployment of AI systems.

Technical Competencies:

• Experience building and managing ML workflows, including model training, deployment, monitoring, and versioning.
• Strong programming skills in Python and familiarity with ML frameworks such as PyTorch or TensorFlow.
• Experience working with cloud platforms such as GCP, AWS, or Azure.
• Strong experience with NLP, speech technologies, conversational AI, or LLM-based applications.
• Experience with multimodal AI, computer vision, or image-based AI workflows is desirable.
• Ability to work collaboratively across technical, product, and field teams.
• Strong communication, documentation, and problem-solving skills.
• Background or experience in agriculture, digital agriculture, or international research environments will be an added advantage.

• Support deployment and optimization of multilingual ASR pipelines for cloud and mobile environments.
• Develop workflows for speech data ingestion, transcription, evaluation, and continuous model improvement.
• Implement automated retraining and fine-tuning pipelines using newly collected field data.
• Support deployment of LLM-based workflows for conversational analysis and trait extraction.
• Monitor model performance, latency, reliability, and drift under field conditions.
• Optimize inference workflows for low-connectivity and resource-constrained environments.
• Support development of multimodal pipelines linking speech, transcripts, metadata, and field images collected through SIKIA.
• Implement workflows for multimodal data validation, synchronization, storage, annotation, and dataset versioning.
• Support training, deployment, and evaluation of multimodal and visual-language AI models.
• Develop scalable workflows for managing multimodal datasets and model outputs across cloud infrastructure.
• Support benchmarking, reproducibility, and optimization of multimodal AI pipelines for field deployment.
• Support development and deployment of AI workflows for disease detection and severity scoring using field images and multimodal data.
• Implement data and evaluation pipelines for disease annotation, validation, benchmarking, and continuous model improvement.
• Support integration of disease scoring workflows within the SIKIA and ONA platforms for field-based data collection and analysis.
• Develop CI/CD pipelines for model training, evaluation, and deployment.
• Manage experiment tracking, model registries, and dataset versioning workflows.
• Implement monitoring and logging across ML services.
• Collaborate with the software development team to integrate RAG pipelines into an agentic deployment structure.
• Support deployment of ML services on GCP, FAIRGrounds, and related infrastructure.
• Support integration of ML services within the SIKIA platform across mobile, backend, API, and cloud systems.
• Ensure compliance with data governance, security, and responsible AI requirements.

• Experience building and managing ML workflows, including model training, deployment, monitoring, and versioning.
• Strong programming skills in Python and familiarity with ML frameworks such as PyTorch or TensorFlow.
• Experience working with cloud platforms such as GCP, AWS, or Azure.
• Strong experience with NLP, speech technologies, conversational AI, or LLM-based applications.
• Experience with multimodal AI, computer vision, or image-based AI workflows is desirable.
• Ability to work collaboratively across technical, product, and field teams.
• Strong communication, documentation, and problem-solving skills.

• Master’s degree in computer science, Data Science, Artificial Intelligence, Software Engineering, or a related field.
• At least 3 years of experience in Machine Learning engineering, MLOps, or deployment of AI systems.