Senior Swift Engineer - AI Platform About the Role
We are building a native application that delivers AI-powered data insights for business users. This is a systems-level contract engagement you'll work across AI runtime, platform integrations, deterministic data execution, or adaptive UI depending on your strengths and interests. The architecture is multi-stage pipeline driven, with strict layered dependencies across 20 Swift Package Manager targets and no shortcuts. The application is a structured, multi-stage pipeline that turns natural language into verified, traceable results. We need engineers who ship real desktop software and take ownership of their technical surface.
What You'll Do - Own one or more technical surfaces within a native AI pipeline: runtime/orchestration, platform/integrations, deterministic execution, or adaptive UI.
- Design and implement production Swift code within a layered SPM architecture (20 targets) with strict dependency boundaries from foundation types through orchestration to application layer.
- Build AI orchestration including structured output extraction, multi-stage routing, tool calling, context management, and response synthesis.
- Implement platform integrations protocol-based connector clients, enterprise data access, entitlements, sandboxing, identity, and persistence boundaries.
- Build deterministic execution pathways joins, aggregations, rankings, filters with full data provenance and traceability.
- Develop adaptive UI using SwiftUI with AppKit interop desktop windowing, keyboard-first workflows, streaming content updates, state management, and accessibility.
- Write and maintain automated tests using XCTest and Swift Testing frameworks.
- Ensure all shared mutable state uses actors or Sendable value types with correct isolation annotations under Swift 6 strict concurrency.
Required Experience - 5 years of professional software engineering experience with Swift
- Experience shipping desktop software (not only)
- Strong understanding of Swift concurrency: actors, async/await, Sendable, structured concurrency, and typed throws
- Experience working within layered, modular SPM architectures with strict dependency management
- Proficiency with protocol-oriented design and dependency injection patterns
- Comfortable with pre-commit hooks, strict linting (SwiftLint, SwiftFormat), and enforced commit conventions
Preferred Experience (one or more of the following) AI Runtime and Orchestration - LLM integration: guided generation, structured output, tool calling, streaming inference - Experience with ML inference frameworks and model orchestration - Token budget management, prompt composition, multistage pipeline orchestration - Evaluation and benchmarking of non-deterministic AI outputs
Platform and Integrations - Protocol-based connector/client architectures (SSE, stdio, REST, OData) - system services: process lifecycle management (launchd, SMJobBless), entitlements, sandboxing - Code signing, notarization, and installer packaging for distribution - Enterprise data access patterns and pagination strategies
Deterministic Execution and Data Systems - TabularData or equivalent data frame processing in Swift - Joins, aggregations, rankings, filters with provenance tracking - Graph or vector retrieval systems, entity resolution - Schema-driven field matching and query planning
Adaptive UI - Production SwiftUI applications with AppKit interop on - Adaptive layouts, multi-panel desktop windowing, keyboard-first navigation - Streaming UI updates tied to async data pipelines - Accessibility compliance and performance optimization (Instruments profiling)
Technical Environment - Swift 6.2, 26 , SPM-only (no Xcode project file)
- 20 SPM targets across foundation, stage, orchestration, and application layers
- LLM integration and ML inference frameworks for AI-powered features
- XCTest and Swift Testing for test automation
- SwiftLint SwiftFormat with strict pre-commit enforcement
- Git-based workflow with conventional commits and CI pipeline integration