Your result
0 / 100
Smaller start: Score your answers against 50 rubrics, $9. 50 AI engineer questions, one rubric each, PDF on Gumroad.
Covered concepts (0)
Missed concepts (0)
Each missed concept is spelled out in the model answer for this question. Read it, then try again out loud.
Get the playbook 40% off with code EARLYBF until Nov 3, 11:59 PM PTEARLYBF: 40% off until Nov 3
Structure
Directional logic
What to improve
The PlaybookWant the full method?
The 0 to 1 AI Engineer Interview Playbook covers the answer frameworks, system-design walkthroughs and an 8-week plan.
Get the playbook 40% off with code EARLYBF until Nov 3, 11:59 PM PTEARLYBF: 40% off until Nov 3Smaller start: Score your answers against 50 rubrics, $9. 50 AI engineer questions, one rubric each, PDF on Gumroad.
Interviewing for PM or TPM? See the full PM package, $59 on Gumroad: a resume system, a PM, TPM and PGM interview handbook, a 1:1 cheat sheet and a PM playbook. PM and TPM content, not AI engineer prep.
想要真人批改？ Want a human review of your answer? Join the waitlist
How the score works
The rubric is the same for every answer and is shown in full here.
| Part | Points | What is checked |
|---|---|---|
| Key concepts | 60 (70 when a question has no logic checks) | Each question has 3 to 10 required concepts taken from its model answer, each with synonyms. Points are the share of concepts your answer mentions. |
| Structure | 20 | A definition or direct answer in the first two sentences (7), how it works or why (7), and an example, number or trade-off (6). |
| Length | 10 | Full points for 80 to 300 words. Fewer points for shorter or much longer answers. |
| Directional logic (IB) | 10 | For investment banking questions where it applies, key phrases are checked for the right direction, for example that net income falls and cash rises when depreciation goes up, or that a higher WACC lowers value. |
Limits: the check matches words and phrases, so it cannot judge whether an explanation is actually correct in every detail, and an answer can score well and still need work. Read the model answer and practice out loud.
Questions you can check
AI Engineer (50)
- What is RAG (retrieval-augmented generation)?
- Explain the transformer architecture
- What is the KV cache and why does it matter for inference?
- Why do LLMs hallucinate, and how do you reduce it?
- What is prompt injection and how do you defend against it?
- Explain LoRA and QLoRA
- RAG vs fine-tuning: when should you use each?
- How do you evaluate an LLM?
- How do you evaluate a RAG system?
- What is an AI agent, and how is it different from a chain?
- What is the difference between pre-training, instruction tuning and RLHF?
- What are LLM scaling laws and how do they change decisions?
- What is tokenization in LLMs and why does it matter?
- What is a context window and how does it affect system design?
- What do temperature and top-p do in an LLM?
- What is LLM quantization and why does it matter for deployment?
- What is Mixture of Experts (MoE) and what are its trade-offs?
- How do reasoning models differ from standard chat models?
- What is the difference between an embedding model and a generative model?
- What is chain-of-thought prompting and when does it help?
- How do you get reliable structured JSON output from an LLM?
- What is function calling (tool calling) in LLMs?
- How do you choose a chunking strategy for RAG?
- What is hybrid search and when should you use it?
- How does re-ranking work in RAG and why use it?
- What is the lost in the middle problem in LLMs?
- How do you choose an embedding model?
- HNSW vs IVF: how do vector indexes trade off recall, speed and memory?
- How do you enforce access control in a RAG system?
- How do you debug poor RAG answer quality?
- What is the ReAct pattern for LLM agents?
- What is the Model Context Protocol (MCP)?
- How do you make AI agents reliable in production?
- How do you handle memory in AI agents?
- Single agent vs multi-agent systems: when is multi-agent worth it?
- Workflows vs agents: how do you decide?
- How do you build a fine-tuning dataset?
- What is catastrophic forgetting and how do you limit it?
- What is model distillation and when would you use it?
- What is DPO (Direct Preference Optimization)?
- How do you make LLM-as-a-judge evaluation reliable?
- How do you A/B test an AI feature?
- What is benchmark contamination and why does it matter?
- Design an LLM customer-support assistant for a SaaS company
- How would you design an LLM gateway for a company?
- How do you design caching for LLM applications?
- How do you reduce LLM inference cost?
- How do you reduce LLM latency?
- What does red-teaming an LLM system involve?
- What are LLM guardrails and where do they sit?
Investment Banking (50)
- Walk me through a DCF
- Walk me through the three financial statements
- If depreciation goes up by $10, walk me through the three statements
- Enterprise value vs equity value: what is the difference?
- What is WACC and how do you calculate it?
- What is terminal value and how do you calculate it?
- Walk me through an LBO model
- How do you do a paper LBO?
- Walk me through a merger model
- Why investment banking?
- Walk me through your resume
- What is your greatest weakness?
- What is the difference between cash and accrual accounting?
- A company prepays $12M of rent for the year: walk me through the three statements
- What is goodwill and how is it created?
- How does a goodwill impairment affect the three statements?
- Walk me through the enterprise value bridge
- Why do you add minority interest (noncontrolling interest) to enterprise value?
- What is deferred revenue and is it a source or use of cash?
- LIFO vs FIFO: how does each affect the financial statements when prices rise?
- What is the difference between unlevered and levered free cash flow?
- How should you treat stock-based compensation in free cash flow and a DCF?
- A company issues $100M of debt to buy back stock: walk me through the three statements
- What is a deferred tax asset vs a deferred tax liability?
- What is working capital and how does an increase affect cash flow?
- What is the difference between capitalizing and expensing a cost?
- How do operating leases appear on the financial statements under ASC 842?
- How do you calculate diluted shares with the treasury stock method?
- How do you calculate cost of equity?
- What beta would you use for a private company?
- Gordon Growth vs exit multiple: which terminal value method should you use?
- Can you do a DCF for a company with negative free cash flow?
- Walk me through how you would build a comparable companies analysis
- Why do precedent transactions usually give higher valuations than comps?
- If you raise the discount rate by 1%, what happens to DCF value?
- Why use NTM multiples instead of LTM multiples?
- What is accretion/dilution in M&A?
- In an all-stock deal, what determines if it is accretive or dilutive?
- What are synergies in M&A?
- Cash vs stock consideration in M&A: what are the trade-offs?
- Walk me through a sell-side M&A process
- What is the difference between an asset deal and a stock deal?
- How do you calculate goodwill in an acquisition (purchase price allocation)?
- What makes a good LBO candidate?
- What are the three ways private equity firms create value?
- What is the difference between IRR and MOIC?
- Which creates more value in an LBO: one turn of multiple expansion or one more turn of leverage?
- What is a dividend recapitalization?
- Why can't you use EV/EBITDA to value a bank?
- Why do you want to work at our bank?
The PlaybookWant the full method?
The 0 to 1 AI Engineer Interview Playbook covers the answer frameworks, system-design walkthroughs and an 8-week plan.
Get the playbook 40% off with code EARLYBF until Nov 3, 11:59 PM PTEARLYBF: 40% off until Nov 3Smaller start: Score your answers against 50 rubrics, $9. 50 AI engineer questions, one rubric each, PDF on Gumroad.
Interviewing for PM or TPM? See the full PM package, $59 on Gumroad: a resume system, a PM, TPM and PGM interview handbook, a 1:1 cheat sheet and a PM playbook. PM and TPM content, not AI engineer prep.
Scores are a self-check for practice, not a prediction of any interview outcome. All 100 investment banking interview questions.