Engineering Manager, iOS · 7+ years in Big Tech
5 years building iOS apps as an engineer, then 2+ years leading the teams that build them. I still review architecture and write code.
Engineering Manager, iOS · Big Tech · years 6 to present
- Run an iOS team end to end: hiring, onboarding, 1:1s, performance reviews, promotions.
- Own the roadmap with product and design, split it into quarters, and keep delivery predictable.
- Set the release process: release trains, feature flags, phased rollouts, crash and hang budgets.
- Grow senior engineers into tech leads and keep architecture decisions with the people who write the code.
- Coordinate with backend, QA and platform teams on contracts, deadlines and incidents.
- Rolled out AI coding tools across the team, with review rules and metrics on what they change.
Senior iOS Engineer · Big Tech · years 1 to 5
- Built and maintained features in a large modular Swift and Objective-C codebase.
- Moved screens from UIKit to SwiftUI and callback code to Swift Concurrency.
- Worked on startup time, scroll performance, memory and crash rate.
- Wrote unit, snapshot and UI tests, and the CI checks that run them.
- Mentored juniors and ran code review for the module I owned.
| Area | Stack |
|---|---|
| Apple platforms | |
| Languages | |
| Tooling | |
| AI tools |
| Practice | What I do with it |
|---|---|
| Agentic development | Daily work in Claude Code, Codex and Gemini CLI. I write custom skills, hooks, subagents and MCP servers so agents follow team conventions. |
| Context engineering | Project instruction files, scoped memory and task decomposition that keep agents on track in large repos. |
| AI in code review and CI | LLM review passes on pull requests, generated test cases, triage of crash reports and flaky tests. |
| Evaluation | Eval suites for prompts and agents, LLM-as-judge with human spot checks, regression runs before a model or prompt change ships. |
| Structured output and tool use | JSON schema outputs, function calling, retries and validation so model responses can feed production code. |
| RAG | Embeddings, chunking, vector search and reranking over docs and codebases. |
| On-device ML | Core ML, Apple Foundation Models, Vision, Speech and Natural Language frameworks, MLX for local models on Apple Silicon. |
| Speech and alignment | ASR pipelines and CTC forced alignment that place transcript words and letters in time. |
| Cost and latency | Prompt caching, model routing by task, token budgets, streaming responses in the UI. |
| Safety and privacy | PII redaction before requests leave the device, guardrails, human approval on destructive agent actions. |
| Team adoption | Guidelines for AI-assisted work, review standards for generated code, measuring impact on cycle time. |

