AI / Machine Learning Engineer

• Build AI features into client products: document understanding, search and retrieval (RAG), classification, forecasting and recommendation.
• Work with large language model APIs and open models - prompt design, evaluation, function calling, guardrails and cost control.
• Build retrieval pipelines: chunking, embeddings, vector stores, re-ranking and honest evaluation of answer quality.
• Fine-tune or adapt models where it genuinely beats prompting, and say so when it does not.
• Deploy models and pipelines behind monitored, versioned APIs with sensible fallbacks.
• Measure real accuracy on real client data, and be transparent with clients about limitations and failure modes.
• Keep current with a field that changes monthly, and filter hype from what actually ships.

• Bachelor's or Master's degree in Computer Science, Data Science, Statistics, Mathematics or a related field.

• 2 to 5 years in machine learning, data science or applied AI engineering.
• Strong Python plus the standard stack: pandas, scikit-learn, PyTorch or TensorFlow.
• Practical LLM application experience - you have shipped something that real users used.
• Experience with vector databases and embedding-based retrieval.
• Ability to take a model from notebook to production service.

• Share a GitHub, paper, Kaggle profile or deployed project with your application.
• You are rigorous about evaluation and resistant to demo-driven development.
• Experience with speech, OCR, computer vision or Nepali-language NLP is a strong plus.

Negotiable - senior band, based on demonstrated ownership and delivery record. Reviewed at six months, with a performance bonus tied to project delivery.

• Competitive salary reviewed annually, plus performance-based increments
• Festival bonus (Dashain/Tihar) and leave encashment as per company policy
• Provident fund, gratuity and insurance in line with the Nepal Labour Act
• Five-day working week with flexible start hours; hybrid options for eligible roles
• Paid annual leave, sick leave and all gazetted public holidays
• Company-funded certifications, courses and conference passes
• Modern hardware, licensed tools and a quiet, well-equipped office in Tokha
• A clear career track with six-monthly reviews - we promote from within first

Please apply only if you meet the core requirements listed above.

Attach an updated CV in PDF format and, where relevant, links to your portfolio, GitHub, Behance, Dribbble or live projects - applications for design and engineering roles without work samples are difficult to assess.

Shortlisted candidates are contacted within seven working days for a short introductory call, followed by a practical task (paid for senior roles) and a final interview with the team lead. Sambara Technologies never charges a fee at any stage of hiring and will never ask for payment for training, equipment or placement.