New opportunity

AI Engineer - Azure Foundry

Latent Bridge · New York, United States

About this role

Latent Bridge seeks an AI Engineer for a contract engagement modernizing an enterprise platform through independently deployable, domain-scoped Micro UIs. The role combines AI-assisted features and agentic workflows on Azure AI Foundry with Angular front ends, .NET backend services, and document and data intelligence. The posting specifies 8–10 years of work experience and calls for candidates who have already delivered comparable projects.

Responsibilities

Design and build AI-assisted features and agentic workflows using Azure AI Foundry.
Build domain-specific Micro UIs in Angular and integrate them with .NET services. Develop pipelines that extract, process, and structure data from unstructured documents, using Databricks and Snowflake for supporting data processing, storage, and retrieval.
Use agentic coding and AI-assisted development as part of delivery, targeting 60%+ AI-generated code with rigorous human review.
Work with architects, business analysts, and domain teams to match each Micro UI to its domain’s access patterns and requirements. Help maintain consistent, secure, maintainable platform engineering standards.

Required experience: Hands-on Azure AI Foundry or directly equivalent Azure AI/ML tooling; .NET backend and Angular front-end development; Databricks and Snowflake; agentic AI development and AI-assisted coding with a production track record of 60%+ AI-generated code; and Micro UI or micro-frontend architectures, ideally with domain-based access segmentation. Candidates must communicate prior project experience clearly in client-facing evaluations.

Preferred experience includes AI-first architecture, document intelligence such as OCR, extraction and classification, platform engineering, business analysis, project or program management on similar engagements, and engineering experience relevant to the client’s industry. The position is listed in New York, United States.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

Azure AI FoundryAzure AI/MLAgentic AI developmentAI-assisted codingAngular.NETDatabricksSnowflakeDocument intelligenceDocument processingOCRData extractionDocument classificationMicro-frontendsPlatform engineeringAI architectureBusiness analysisProject managementCommunication

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Deliver AI features on Azure AI Foundry
The role designs and builds AI-assisted features and agentic workflows on Azure AI Foundry. Prepare a concrete account of a comparable delivery, including the workflow, your engineering decisions, and what reached production.
Build domain-scoped Micro UIs
The engagement centers on independently deployable Angular Micro UIs integrated with .NET backend services. Be ready to explain how a prior micro-frontend was scoped, integrated, and delivered.
Turn documents into usable data
The posting calls for pipelines that extract, process, and structure unstructured documents, with Databricks and Snowflake supporting data processing, storage, and retrieval for AI features. Prepare to describe an end-to-end pipeline and its data handoffs.
Use AI-assisted coding with human review
AI-assisted development is a core delivery practice, with a target of 60%+ AI-generated code and rigorous human review. Prepare to explain how you measured that contribution and reviewed production code, using a real example if available.
Align delivery across domains
The engineer works with architects, business analysts, and domain teams on access patterns and requirements, while contributing to secure, consistent, maintainable platform standards. The posting also emphasizes clearly explaining relevant past work in client-facing evaluation.

Likely challenges to think through

Balance domain autonomy with shared standards
Preparation hypothesis: independently deployable, domain-scoped UIs may need different access patterns without diverging on security or maintainability. Prepare an example of how you defined domain boundaries while preserving shared engineering conventions.
Keep review rigorous at high AI-code volume
Preparation hypothesis: the 60%+ AI-generated-code target could increase delivery speed while making defects or unclear ownership harder to catch. Prepare to discuss specific human review, testing, and correction practices you used—not just the generation percentage.
Connect imperfect documents to reliable AI features
Preparation hypothesis: extracting structured data from unstructured documents and making it retrievable through data platforms may expose quality and integration trade-offs. Prepare to explain how you checked extraction results and handled data handoffs in a comparable project.

Practice your answers

Technical practice

How would you design an AI-assisted feature and agentic workflow using Azure AI Foundry?

What to coverThe role involves building AI-assisted features and agentic workflows on Azure AI Foundry. Preparation suggestion: explain the workflow’s boundaries, its integration with application services, and how you would evaluate whether it meets the domain’s needs.

How would you troubleshoot an agentic workflow that produces inconsistent results?

What to coverAgentic AI development is a required part of the role. Preparation suggestion: discuss how you would reproduce the issue, inspect the workflow’s steps and inputs, and decide whether to revise the workflow or add human review.

How would you divide an enterprise interface into independently deployable, domain-scoped Micro UIs?

What to coverThe initiative centers on independently deployable Micro UIs built with Angular. Preparation suggestion: explain domain boundaries, integration points, and the trade-off between local autonomy and a consistent user experience.

How would an Angular Micro UI integrate with its .NET backend service?

What to coverThe posting calls for Angular front ends integrated with .NET services. Preparation suggestion: describe the interface between them, how domain-specific behavior would be kept clear, and how you would diagnose an integration failure.

How would you handle domain-based access patterns across Micro UIs?

What to coverThe role calls for aligning each Micro UI with its domain’s access patterns; domain-based access segmentation is also mentioned. Preparation suggestion: explain where access decisions should be enforced and how you would check that UI behavior and backend access remain consistent.

How would you turn unstructured documents into usable structured data?

What to coverImplementing document intelligence pipelines is a stated responsibility, while OCR, extraction, and classification are preferred experience. Preparation suggestion: walk through processing stages, quality checks, and how you would investigate incorrect or missing extracted fields.

How would you use Databricks and Snowflake to support an AI feature?

What to coverThe posting names both platforms for backend processing, storage, and retrieval. Preparation suggestion: distinguish the work you would assign to each in a relevant design, and explain the data-flow and maintenance trade-offs without assuming a prescribed architecture.

How would you review AI-generated code before production delivery?

What to coverThe delivery workflow targets 60%+ AI-generated code with rigorous human review. Preparation suggestion: explain how you would verify correctness, security, maintainability, and integration rather than treating generated code as inherently ready to ship.

Which engineering standards should be shared across domain Micro UIs?

What to coverContributing to consistent, secure, maintainable platform standards is part of the role; shared services and golden paths appear as preferred experience. Preparation suggestion: discuss what should be standardized and where domain teams should retain flexibility.

When would an AI-first architecture be preferable to a narrower AI-assisted feature?

What to coverAI-first system architecture is listed as preferred experience, alongside the responsibility to build AI-assisted features. Preparation suggestion: compare the value and complexity of each approach for a domain workflow, including how you would judge whether AI is necessary.

Behavioural practice

Tell me about a time you reconciled different domain and architecture requirements.

What to coverThe role involves working with architects, business analysts, and domain teams. Preparation suggestion: use a real situation, your actions to resolve the differences, and the resulting decision or outcome.

Describe a time your review changed or rejected AI-generated code.

What to coverThe posting pairs extensive AI-assisted coding with rigorous human review. Preparation suggestion: give a real example of what you found, what you did, and the result; avoid claiming a code-generation percentage you cannot substantiate.

Tell me about a document-processing quality problem you helped resolve.

What to coverBackend document intelligence is central to the role. Preparation suggestion: describe the situation, your investigation and corrective action, and the observed result, using a real example if you have one.

Describe a time you balanced a shared platform standard with a domain team’s needs.

What to coverThe role combines domain-scoped Micro UIs with consistent platform engineering standards. Preparation suggestion: explain the competing needs, your contribution to the decision, and what resulted.

How have you explained a comparable technical delivery to a client-facing audience?

What to coverThe posting calls for clear communication of prior relevant project experience during client-facing evaluation. Preparation suggestion: choose a real situation, explain how you communicated your actions and technical choices, and state the outcome without overstating your role.

Sharpen your knowledge

Topics to prepare

Azure AI Foundry and agentic workflow design
The role calls for AI-assisted features and agentic workflows on Azure AI Foundry. As a revision exercise, sketch a workflow from user request to agent action to human review, and explain where errors could occur.
Domain-scoped Micro UI architecture
The posting describes independently deployable Angular front ends aligned to domain access patterns. Draw boundaries for two hypothetical domains and explain how their UIs would integrate with .NET services without blurring access rules.
Document intelligence pipeline design
The role includes extracting, processing, and structuring unstructured documents; OCR and classification are preferred experience. Trace a sample document through extraction to structured output, marking points where quality checks would help.
Databricks and Snowflake data flow
Both platforms are named for processing, storage, and retrieval supporting AI features. Diagram a plausible division of responsibilities between them, labeling it as a design exercise rather than the employer’s prescribed architecture.
AI-assisted delivery and review
The posting targets 60%+ AI-generated code in production delivery with rigorous human review. Revisit how you would assess generated code for correctness, security, and maintainability, and explain what evidence would support a production-delivery claim.

Skills to revise

Build a small Foundry workflow
Practice implementing an AI-assisted feature or agentic workflow in Azure AI Foundry, or directly equivalent Azure AI/ML tooling, then walk through its behavior and limitations. Hands-on experience is required.
Connect an Angular Micro UI to .NET
Build a small domain-scoped Angular view backed by a .NET service; exercise a domain access boundary and review the integration. Both technologies and Micro UI delivery are explicit requirements.
Process a document into usable data
Practice extracting fields from an unstructured document, checking the result, and explaining how Databricks and Snowflake could support processing and retrieval. This mirrors the stated document and data responsibilities; the sample implementation is a preparation exercise.
Review AI-generated code
Take a small AI-assisted code change, inspect and correct it, and explain what human review caught. The posting explicitly pairs agentic coding with rigorous review; the exercise does not substitute for its requested production track record.
Explain a relevant delivery example
Rehearse a concise account of a real project you have delivered, if applicable: your contribution, architecture, trade-offs, and outcome. The posting calls for clear discussion of prior work in client-facing evaluations; do not claim experience you cannot substantiate.

Your preparation checklist

Select a comparable project walkthrough
The posting asks for demonstrable delivery experience and clear explanations in client-facing evaluations. Prepare a concise account of a relevant project, your decisions, production outcome, and what you would change.
Sketch a domain-scoped Micro UI design
As interview practice, diagram an independently deployable Angular Micro UI integrated with .NET services. Explain domain access boundaries, deployment ownership, and how shared standards keep UIs secure and maintainable.
Walk through an Azure AI Foundry workflow
Prepare an example of an AI-assisted feature or agentic workflow using Azure AI Foundry or directly equivalent Azure AI/ML tooling. Identify the agent’s tasks, service integrations, and points requiring human oversight.
Map a document-to-data pipeline
Draw how unstructured documents become structured data and how Databricks and Snowflake support processing, storage, and retrieval. Discuss OCR or classification only if relevant to your experience; these are preferred specializations.
Document AI-assisted coding practices
The posting targets 60%+ AI-generated code in production with rigorous human review. If you have verifiable examples, prepare how you measured that share and reviewed, tested, and corrected generated code; do not claim a figure you cannot substantiate.
Prepare a domain-requirements discussion
Rehearse how you would work with architects, business analysts, and domain teams to turn access patterns and requirements into a Micro UI specification. Bring one concrete example of a trade-off to discuss.
Write focused questions for the client
Ask which domains and document types are in scope, how Micro UIs will be deployed and governed, and how the team evaluates AI workflows and code review. These are questions to resolve, not established details of the engagement.

Availability last checked: Nov 4, 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in United States. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…