AI/ML Solution Architect
New
J
JobgetherAI/ML
Based in India, Comfort working across international time zones; early mornings or late evenings occasionally.Full-TimeSenior
Salary not disclosed
Apply NowOpens the employer's application page
Job Details
- Languages
- Excellent written English
- Experience
- 8+ years of experience building software, data, or AI/ML systems
- Required Skills
- AWSPythonMachine LearningAzure
Requirements
- Have 8+ years of experience building software, data, or AI/ML systems, including recent experience as a Solution Architect, AI/ML Architect, Principal Engineer, or Staff Engineer with end-to-end architecture ownership.
- Have shipped at least one production AI/ML or GenAI system at scale that passed security, cost, and operational reviews.
- Create delivery-ready architecture blueprints covering services, data stores, IAM, trust boundaries, failure modes, evaluation strategies, and cost models.
- Bring deep Azure and/or AWS experience, including landing zones, private-endpoint architectures, cloud security controls, and infrastructure and inference costs.
- Have strong Python skills and practical understanding of classical machine learning, model evaluation, LLM APIs, embeddings, vector search, and production AI behavior.
- Have hands-on LLMOps/MLOps experience, including versioning, evaluation harnesses, CI/CD release gates, observability, experiment tracking, and rollback strategies.
- Understand RAG over structured knowledge, hybrid search, document AI/OCR, multilingual AI, agent security, prompt-injection risks, and AI evaluation methodologies.
- Have experience working with non-technical stakeholders and facilitating workshops that translate business, programme, or government requirements into technical solutions.
- Have excellent written English and strong documentation skills for creating clear technical contracts for engineering teams.
- Be able to work across international time zones, with occasional early mornings or late evenings.
Responsibilities
- Lead discovery workshops with programme, government, domain, and engineering stakeholders, translating requirements into technical architectures.
- Own solution design from discovery through production, producing service breakdowns, data flows, trust boundaries, diagrams, decision records, cost models, and implementation blueprints.
- Architect production AI pipelines for generation, RAG, knowledge-graph retrieval, multi-agent workflows, LLM-as-judge systems, document AI/OCR, and human review.
- Define AI system contracts covering schemas, model and prompt versions, token usage, latency, evaluation criteria, and failure handling.
- Design secure human-in-the-loop workflows with audit trails, configuration controls, de-identification, and protection of sensitive information.
- Own Azure and/or AWS reference architectures covering identity, private networking, storage, vector search, monitoring, observability, LLMOps/MLOps, and production operations.
- Establish evaluation frameworks and release gates to prevent underperforming models, prompts, or configuration changes from reaching production.
- Apply FinOps principles through token budgets, inference and OCR cost controls, model routing, fallbacks, and degradation strategies.
- Remain accountable through implementation and the first production cycle, ensuring architectures are practical, resilient, secure, and operationally viable.
- Produce reusable architecture patterns, technical documentation, and country-adaptation frameworks, and mentor technical leads on AI-driven versus deterministic workflows.
View Full Description & ApplyYou'll be redirected to the employer's site
