American Express AI Engineer Interview Questions

American Express·AI Engineer·Technical Phone Screen

Jun 2026

Technical interview for an AI Engineer role at American Express focused entirely on how you evaluate LLM and agent systems in production. One big meaty question that branched into a lot of sub-topics. Dense but interesting.

How would you design an end-to-end evaluation framework for an AI system that uses LLMs or agents? Walk through both offline and online approaches, how you handle agent-specific metrics, regression testing across model versions, and where human eval fits in. Also discuss what can go wrong with these setups.

“This was basically the whole interview packed into one question.”

View Post

American Express·AI Engineer·Onsite - System Design / Architecture

Jun 2026

System design round at Amex for an AI Engineer position, focused almost entirely on fault tolerance and reliability for AI systems. Pretty deep dive, more infrastructure-heavy than I expected for an AI role.

How would you design a fault-tolerant AI system that can recover from failures, covering both stateless and stateful components?
How do you handle idempotency and exactly-once semantics when retrying LLM or tool calls?
What retry and circuit breaker strategies would you use around LLM and retrieval backend calls?
How would you design graceful degradation when a downstream model or retrieval backend becomes unhealthy?
How would you approach multi-region replication and failover for an AI system, including both model artifacts and runtime state?
How do you handle KV-cache invalidation, prompt versioning, and reproducibility when model versions change?

“This was the anchor question and it sprawled into everything else.”

View Post

American Express·AI Engineer·Technical Phone Screen

Jun 2026

Had a deep-dive technical screen at Amex for an AI Engineer role, basically one long question about how memory works in AI agents. Pretty niche topic and the interviewer clearly knew their stuff, so there was nowhere to hide.

How does short-term or working memory function in an AI agent, and what are its limitations?
How would you design long-term memory for an agent that needs to persist information across sessions?
What is the difference between episodic and semantic memory in the context of an AI agent, and how would you implement each?
Walk me through different retrieval strategies for agent memory and when you'd pick one over another.
How would you handle forgetting or compaction in an agent's memory to keep it from growing unbounded?
How would you design the read and write APIs for an agent memory system?
How do you handle consistency and conflicts when multiple agents or sessions share the same memory store?
How would you evaluate whether an agent's memory is actually working well in production?

“I talked about the context window being the agent's active workspace and how everything outside it is effectively invisible unless you retrieve it.”

View Post

American Express·AI Engineer·Onsite - System Design / Architecture

Jun 2026

System design round at Amex for an AI Engineer role, focused entirely on long-running multi-agent systems. Pretty deep technically, more so than I expected for a financial services company. Left feeling like I'd covered maybe 60% of what they were probing for.

How would you design a multi-agent system that needs to run reliably for hours or days, handling crashes and restarts without losing progress?
How do you ensure tool calls made by agents are idempotent, and what retry and recovery semantics do you apply when something fails mid-execution?
Describe how you would implement supervision in a multi-agent system, where a parent agent monitors and can restart child agents that fail or misbehave.
What communication patterns would you use between cooperating agents, and what are the tradeoffs between message passing versus shared memory or a blackboard architecture?
How do you approach debugging and observability for agent systems that produce very long execution traces spanning many steps and possibly multiple agents?
How do you manage LLM API costs and rate limits for agents running over long time horizons, especially when multiple agents are running concurrently?
How would you design human-in-the-loop interrupt points in a long-running agent workflow, where a human needs to review or approve something before the agent continues?
What failure modes are specific to long-running agents, such as drift, looping, or runaway costs, and how do you mitigate them?

“This is where I spent most of my time and still felt like I left things on the table.”

View Post

American Express·AI Engineer·Technical Phone Screen

Jun 2026

Technical screen for an AI Engineer role at Amex focused almost entirely on how agents extend beyond what the base LLM can do. Pretty deep dive, more systems-thinking than I expected for a phone screen.

How does tool use and function calling work in AI agents? Walk through schema design, how the agent routes to the right tool, and what happens when a tool call fails.
How do plug-in style or MCP-style integrations extend an agent's capabilities, and what are the security and reliability trade-offs involved?
When would you use code execution as a tool for an agent, and what sandboxing concerns come with it?
How does retrieval function as a capability for an AI agent, and how does it differ from just stuffing more context into the prompt?
Explain the planner/executor pattern in agent architectures. What are the benefits and failure modes?
How does an agent discover and select the right tools at runtime, especially when the tool set is large?
How do you decide whether a new capability should be added as an external tool versus fine-tuned or baked into the model itself?

“I started with the schema side and felt pretty solid there, JSON schemas, typed parameters, descriptions that actually help the model pick correctly.”

View Post

American Express·AI Engineer·Technical Phone Screen

Jun 2026

Amex AI Engineer interview that went deep on LLM security, specifically prompt injection. The whole session felt like one long threat modeling exercise and I was not fully prepared for how far into the weeds they wanted to go on mitigations.

What is the difference between direct prompt injection and indirect prompt injection, and why does the distinction matter for threat modeling?
Walk through the main threat categories in LLM agent systems, covering data exfiltration, privilege escalation, jailbreaks, and tool misuse.
What mitigation strategies would you apply to defend an LLM agent against prompt injection, and what are the limits of each?
How would you design an end-to-end defense for an LLM agent system deployed in a financial services context, assuming defense-in-depth?
How does content provenance help defend against indirect prompt injection, and where does it fall short?

“I explained direct as a user crafting malicious input themselves vs indirect as the attack arriving through retrieved content or tool output.”

View Post

American Express·AI Engineer·Technical Phone Screen

May 2026

Technical screen for an AI Engineer role at Amex. One meaty conceptual question about how LLMs, RAG, and agents relate to each other. Felt like they wanted to see if you actually understood the architecture or were just name-dropping buzzwords.

Walk me through the differences between a base LLM, a RAG system, and an autonomous agent. How do they relate to each other, when would you use each, and how do they compose together?

“This question is deceptively wide.”

View Post

American Express·AI Engineer·Onsite - System Design / Architecture

May 2026

System design round at Amex for an AI Engineer role, focused almost entirely on concurrency in multi-user agent systems. Pretty deep technically and covered a lot of ground in one question.

How would you design a multi-user AI agent system to handle concurrency safely? Walk through context isolation per user, shared resource management (model servers, vector stores, tool backends), fair queueing, request batching vs. latency SLAs, session memory isolation, and consistency when multiple sessions write to shared agent state. Also cover observability and how you'd test for concurrency bugs.

“This was a lot to unpack in one question.”

View Post

American Express·AI Engineer·Onsite - System Design / Architecture

May 2026

Thirty-minute system design case at Amex for an AI Engineer role. The whole thing was a deep dive on building a RAG pipeline end to end, from raw document ingestion all the way through evaluation and ops. Dense but interesting.

Design a Retrieval-Augmented Generation system over a corporate document corpus. Walk through the full pipeline from ingestion to generation, including retrieval strategy, evaluation, and operational concerns.
How would you handle multi-tenant access control when users should only retrieve documents they're authorized to see?
What chunking strategy would you use and why? How does chunk size affect retrieval quality?
How would you evaluate the RAG system, specifically around faithfulness and answer quality?
How would you keep the index fresh as documents are updated or deleted in the corpus?

“This is a big open-ended one and I think I frontloaded too much time on chunking strategy and not enough on the access control piece, which came up later as a follow-up and I was a bit scrambled.”

View Post

American Express·AI Engineer·Technical Phone Screen

Apr 2026

Short Karat-style debugging round for an AI Engineer role at American Express. Two parts: find and fix a bug in existing Python code by reading failing test output, then write a new function to make more tests pass. The whole thing was 20 minutes so speed of diagnosis mattered more than anything else.

You're given a set of Python classes and a failing test suite. Read the test output, identify which function contains the bug, explain why the test is failing, and fix it with a minimal change.
Given a short spec, write a new function from scratch so that a set of currently-failing tests will pass.

“The tricky part isn't writing code, it's resisting the urge to rewrite things.”

View Post