Machine Learning & Operations (MLOps) Consultancy – NDIZI Project

Organization: Alliance of Bioversity International and CIAT
Project: NDIZI (NLP to Develop and Innovate Zero-shot Intelligence)
Platform: SIKIA AI (voice-first multimodal data system)
Duty Station: Arusha, Tanzania (also eligible Kenya-based applicants)
Duration: 11 months (full-time consultancy)

---

About the Organization

The Alliance of Bioversity International and CIAT is part of CGIAR and focuses on:

• Agricultural biodiversity research
• Sustainable food systems transformation
• Climate change adaptation solutions
• Biodiversity conservation
• Nutrition and food security

The organization develops science-driven solutions to address global challenges such as:

• Malnutrition
• Climate change
• Biodiversity loss
• Environmental degradation

---

Project Background (NDIZI & SIKIA)

The NDIZI project focuses on building advanced AI systems to support agricultural research through the SIKIA platform, a:

• Voice-first AI system
• Multimodal data collection platform
• Conversational analytics engine

Core Use Cases

• Understanding farmer preferences
• Disease detection and crop health scoring
• Climate adaptation and environmental response

The system connects:

• Controlled agricultural trials
• Real-world farm environments
• AI-driven field data collection

---

Role Summary

The MLOps Consultant will support end-to-end machine learning operations for:

• Speech AI systems
• NLP pipelines
• Multimodal AI workflows
• Disease detection models
• Field-deployable AI systems

The role focuses on moving models from research prototypes into production-ready, scalable systems used in real agricultural field environments.

---

Key Responsibilities

1. Speech & NLP Systems

• Deploy multilingual ASR (Automatic Speech Recognition) systems.
• Build speech data ingestion and transcription pipelines.
• Implement continuous model retraining using field data.
• Deploy LLM-based conversational analysis workflows.
• Monitor performance, drift, latency, and reliability.
• Optimize models for low-connectivity environments.

---

2. Multimodal AI Pipeline Development

• Build pipelines integrating:
  • Speech
  • Text transcripts
  • Metadata
  • Field images
• Manage dataset versioning and annotation workflows.
• Train and deploy multimodal AI and vision-language models.
• Ensure reproducibility and benchmarking of AI systems.
• Optimize storage and synchronization of multimodal datasets.

---

3. Disease Detection & Scoring Systems

• Develop AI workflows for crop disease detection.
• Build severity scoring models using field images.
• Implement annotation and validation pipelines.
• Integrate outputs into SIKIA and ONA platforms.
• Improve model accuracy through continuous feedback loops.

---

4. MLOps Infrastructure & System Integration

• Build CI/CD pipelines for ML model lifecycle.
• Implement:
  • Experiment tracking
  • Model registries
  • Dataset versioning systems
• Set up monitoring and logging systems.
• Deploy ML systems on cloud infrastructure (e.g. GCP).
• Integrate ML services into SIKIA platform APIs and mobile apps.
• Ensure responsible AI and data governance compliance.

---

Deliverables

Deliverable 1 – Inception Report

Timeline: Month 1

• Technical strategy
• Architecture plan
• Deployment roadmap
• 11-month work plan

💰 12,000,000 TZS

---

Deliverable 2 – MLOps Infrastructure Setup

Timeline: Month 5

• CI/CD pipelines
• Model registry setup
• Dataset versioning system
• Initial ML deployment workflows
• Model development support

💰 18,000,000 TZS

---

Deliverable 3 – System Integration

Timeline: Month 8

• Integration of ML systems with SIKIA platform
• Deployment across mobile and backend systems
• Speech, multimodal, and disease pipelines operational

💰 22,000,000 TZS

---

Deliverable 4 – Monitoring & Optimization

Timeline: Month 10

• Model monitoring systems
• Drift detection
• Performance benchmarking
• Optimization recommendations

💰 16,000,000 TZS

---

Deliverable 5 – Final Report & Handover

Timeline: Month 11

• Final technical documentation
• System handover package
• Deployment guides
• Knowledge transfer materials

💰 22,000,000 TZS

---

Total Consultancy Value

💰 90,000,000 TZS (approx.)

---

Required Education

Applicants must hold a Master’s degree in:

• Computer Science
• Data Science
• Artificial Intelligence
• Software Engineering
• Related field

---

Required Experience

Minimum 3+ years in:

• Machine Learning Engineering
• MLOps
• AI system deployment

---

Technical Skills Required

Core ML Engineering

• Python programming (strong proficiency)
• ML pipelines (training, deployment, monitoring)
• Model versioning and lifecycle management

Frameworks & Tools

• PyTorch or TensorFlow
• Cloud platforms (GCP, AWS, Azure)
• CI/CD systems for ML

AI Specializations

• NLP and speech processing
• LLM-based systems
• Conversational AI

Desirable Skills

• Computer vision / multimodal AI
• Agriculture or digital agriculture experience
• Research environment exposure (CGIAR or similar)

---

Key Competencies

• Strong MLOps architecture design
• End-to-end ML system deployment
• Cross-functional collaboration (engineering + field teams)
• System monitoring and optimization
• Strong documentation and communication skills
• Problem-solving in production AI environments

---

Ideal Candidate Profile

This role is best suited for professionals who:

✔ Have strong ML engineering + DevOps hybrid experience
✔ Can move models from research → production systems
✔ Understand cloud-based AI infrastructure
✔ Have worked with NLP, speech, or multimodal AI systems
✔ Are comfortable working in field-based research environments
✔ Can build scalable AI pipelines for real-world deployment

---

Strategic Importance of the Role

This consultancy directly supports:

• Agricultural AI innovation in Africa
• Real-world deployment of speech + vision AI systems
• Climate-resilient crop research
• Farmer-centered digital data collection systems
• Scaling research AI into production infrastructure

---

If you want, I can also:

• Compare this role with the others you shared (which one fits you best)
• Break down required skills vs your CV
• Or help you write an application/cover letter for it

CLICK HERE TO APPLY

Applications should include:

• CV.
• Technical proposal.
• Financial proposal.

All documents should be saved as one document using the candidate’s last name and first name for ease of sorting.

Email applications will not be considered.

Only short-listed candidates will be contacted.

Important recruitment notice

The Alliance does not charge a fee at any stage of the recruitment process, including application, interview meeting, processing, or training.

The Alliance also does not concern itself with information on applicants’ bank accounts.

Application closing date

Applications closing date: 21 June 2026