Senior Software Engineer (R&D – AI)

About Us

Mauritius Network Services Limited was incorporated in 1994 on a public private partnership model and has evolved over the years to be a trusted partner to Government, Businesses and Citizens for trade and business facilitation.

Our purpose is to implement, operate and maintain digital systems and services to enable secure, cost-effective and reliable interactions with government agencies, businesses and citizens, anywhere, anytime.

About Us

Mauritius Network Services Limited was incorporated in 1994 on a public private partnership model and has evolved over the years to be a trusted partner to Government, Businesses and Citizens for trade and business facilitation.

Our purpose is to implement, operate and maintain digital systems and services to enable secure, cost-effective and reliable interactions with government agencies, businesses and citizens, anywhere, anytime.

• Deadline Date: 04/09/2026
• Purpose Of The Job

As a Senior Software Engineer within the Research & Development function, you will explore, prototype and prove out emerging technologies — with a strong focus on Artificial Intelligence and Generative AI — and turn the most promising of them into production-ready capabilities across the company’s project portfolio.

You will spend a significant part of your time building Proofs of Concept (Po C): framing the problem, selecting the right approach, building a working prototype quickly, and evaluating it objectively against business and technical criteria.

Once a Po C is validated, you will act as the technical reference for its adoption, guiding delivery teams through implementation in real projects, establishing reusable patterns, accelerators and standards, and mentoring engineers along the way.

The role calls for an inventive, curious and creative mind: someone who challenges the way things are currently done, is comfortable with ambiguity and rapid iteration, and can distinguish genuine value from hype.

Main Responsibilities

Research, Ideation and Technology Watch

• Maintain an active watch on AI/Gen AI research, models, frameworks, tooling and vendor offerings, and translate findings into opportunities relevant to the company’s products, clients and internal ways of working.
• Generate, challenge and refine ideas for new AI-enabled capabilities, features and services; think laterally about how existing business problems can be reframed and solved differently.
• Benchmark and compare technology options (models, frameworks, vector stores, platforms, licensing and hosting models) and make evidence-based recommendations.
• Assess feasibility, cost, performance, and risk early, and be equally willing to recommend stopping a line of investigation as pursuing it.
• Proof Of Concept Design And Delivery
• Own Po Cs end to end: define the hypothesis, success criteria and evaluation method before building.
• Build working prototypes rapidly, prioritising learning speed while keeping code clean enough to be understood, extended and handed over.
• Design and run objective evaluations (accuracy, relevance, latency, cost per request, hallucination and failure modes, human review) rather than relying on demo-quality impressions.
• Produce clear Po C outcome reports covering results, limitations, risks, estimated effort to productionise, and a go / no-go recommendation.
• Demonstrate results to technical and non-technical audiences, including management, business stakeholders and clients.
• Applied AI And Generative AI Engineering
• Design and build LLM-based solutions: retrieval-augmented generation (RAG), agentic workflows, tool/function calling, structured output, and orchestration across multiple services.
• Apply advanced prompt engineering, context engineering and, where justified, fine-tuning or model adaptation techniques.
• Work with embeddings, vector databases, chunking and indexing strategies, semantic search and re-ranking.
• Integrate AI capabilities into existing enterprise architectures — Java / Spring Boot backends, REST APIs, messaging, relational and No SQL data stores — rather than treating them as isolated experiments.
• Build evaluation, observability and guardrail layers around AI features (tracing, prompt/response logging, regression test suites, cost and token monitoring).
• Explore and apply classical machine learning and data-driven techniques where they are a better fit than generative approaches.
• Industrialisation And Transition To Delivery Teams
• Convert validated Po Cs into reusable assets: reference implementations, shared libraries, templates, accelerators and architectural blueprints.
• Define the standards, patterns and guardrails that project teams must follow when implementing AI capabilities.
• Work alongside project teams during the first implementations — pairing, reviewing code and design, and unblocking technical issues — until the team is autonomous.
• Ensure solutions moving into projects are production-grade: secure, observable, testable, scalable and maintainable within the company’s CI/CD and deployment practices.
• Technical Guidance, Mentoring And Enablement
• Act as the technical reference on AI topics for engineers, architects, QA and project managers across the organisation.
• Mentor and upskill engineers through pairing, code reviews, internal workshops, brown-bag sessions and hands‑on labs.
• Promote effective and responsible use of AI‑assisted engineering tools within delivery teams to improve productivity and quality.
• Contribute to technical assessments, solution design reviews and effort estimation for AI‑related opportunities and proposals.
• Responsible AI, Security And Governance
• Apply data protection and confidentiality requirements to every experiment, including client data handling, anonymisation and data residency considerations.
• Identify and mitigate AI‑specific risks: prompt injection, data leakage, insecure tool execution, model and supply‑chain risk, bias, and over‑reliance on model output.
• Ensure human‑in‑the‑loop controls, traceability and explainability are designed in where decisions carry business or regulatory impact.
• Contribute to internal AI usage policies, review checklists and governance practices.
• Documentation, Communication And Collaboration
• Produce clear technical documentation: architecture and sequence diagrams, API specifications, evaluation results, decision records and adoption guides.
• Maintain an accessible knowledge base of experiments carried out, including those that did not succeed and the reasons why.
• Collaborate closely with architects, backend and frontend engineers, data and BI teams, Dev Ops, QA and product stakeholders.
• Engage with clients and partners when required to present capabilities, gather requirements and shape AI‑related opportunities.
• Participate actively in Agile ceremonies and contribute to R&D roadmap and backlog definition.

Qualifications And Experience

• Degree in Computer Science, Software Engineering, IT or equivalent.
• Minimum 5 years of professional software engineering experience, with a strong track record of delivering production systems.
• Demonstrable hands‑on experience building applications on top of large language models (RAG, agents, tool calling, prompt engineering) — through professional work, side projects or open‑source contributions.
• Strong backend engineering skills, ideally in Java and the Spring ecosystem; proficiency in Python for AI/ML work.
• Solid understanding of OOP, design patterns, software architecture, REST API design and microservices.
• Experience with relational and No SQL databases, and exposure to vector databases (e. g. pgvector, Qdrant, Elasticsearch).
• Experience with Git, Maven, CI/CD pipelines, Docker and containerised deployment; Kubernetes is an advantage.
• Familiarity with AI/LLM frameworks and platforms (e. g.

Spring AI, Lang Chain / Lang Graph, Llama Index, Hugging Face, Model Context Protocol) and with major model providers and open‑weight models.

• Understanding of AI evaluation methods, and of the cost, latency and reliability trade‑offs of different architectures.
• Nice to have: classical ML/deep learning background, MLOps tooling, cloud AI services (AWS / Azure / GCP), data engineering experience, GPU and model‑serving experience.

Skills And Personal Attributes

• Genuinely innovative and creative: naturally questions the status quo, proposes original approaches, and enjoys open‑ended problems.
• Strong experimental mindset — forms hypotheses, tests them quickly, measures results, and draws honest conclusions.
• Comfortable with ambiguity, fast iteration and the possibility that an experiment fails; treats negative results as valuable.
• Pragmatic judgement: able to separate genuine business value from technology hype and to know when a simpler solution is the right one.
• Excellent communicator, able to explain complex technical concepts clearly to engineers, management and clients.
• Natural mentor and influencer who builds capability in others rather than concentrating knowledge.
• Self‑driven, autonomous and continuously learning, with strong ownership and delivery discipline.

Mauritius Network Services Ltd reserves the right to change, update, or withdraw any job vacancy without prior notice.

Posting a position on this website does not guarantee employment.

Only candidates selected for further consideration will be contacted.

#J-18808-Ljbffr