Your preparation is real. The gap is not knowledge — it is not having practised explaining, defending, and troubleshooting under pressure, with someone who pushes back. Sharpen is where you close it.
The real problem
You can run the lab, read the RFC, and reason through a fault. But knowing something and performing it under pressure — to a bar-raiser, in real time, with follow-up questions — are two different skills. One comes from studying. The other only comes from reps against something that pushes back.
The senior engineer
You have shipped real infrastructure. You know what a BGP route leak looks like and how to contain it. But when the design review pauses and the room turns to you, the words don't come as fast as they did. Depth that isn't exercised goes cold — and cold depth does not show up when it counts.
The candidate
You have done the labs, studied the docs, and you understand the material. But there is a gap between knowing a protocol and explaining it clearly to someone who will keep asking until the edge cases surface. That gap does not close by studying more. It closes by practising under pressure.
What's inside
Not a course. Not flashcards alone. A practice studio with six modes — each designed to surface a gap you didn't know you had.
Mock Interview Studio
One question, then 2–3 bar-raiser-style follow-ups based on what you actually said. Skip a trade-off and it asks about it. At the end: a breakdown by technical accuracy, depth, structure, and delivery. The closest thing to a real panel you can get without booking one.
Meeting Simulator
You join a design review already in progress. AI personas push back, ask for justification, and wait for your recommendation. Practice the moment that matters most — the one where the room goes quiet and everyone looks at you.
Troubleshooting Labs
Branching incidents with realistic BGP tables, OSPF neighbour output, and traceroutes. Work through the fault, state your hypotheses, narrow the fault domain — exactly like a production incident, with a score at the end.
Speech Feedback
Every voice answer is scored for filler words per minute, hedging language, pace, and trailing endings. Numbers on the habits most engineers ignore until they are on camera in front of a panel — and it is too late.
Daily Depth Drill
One AI-generated question calibrated to your level, every day. Answer by voice or text, get a scored breakdown and a model answer. Flashcards are built from your gaps and scheduled by SM-2 spaced repetition so you review before you forget.
Routing Lab
A BGP table or OSPF neighbour output with 2–3 planted faults. Name each error, explain the root cause, and propose the fix. Scored on diagnostic accuracy, reasoning quality, and impact assessment.
Who it's for
What you have lacked until now is not the knowledge. It is the reps.
Years of production experience, but the interview muscle has gone soft. You know the answers — you just need to get them out of your head quickly, clearly, and with conviction. Sharpen gives you a place to practise at the level you actually work at, not a simplified version of it.
Strong fundamentals, lab time, and real motivation. The gap is converting knowledge into performance under pressure and building the vocabulary to explain it to someone who does not have your notes in front of them. Every drill builds the muscle memory for that moment.
Topics covered
Questions are generated fresh each session from these domains, calibrated to your level — so you practise the same material from different angles until the explanation is automatic.
How it works
Choose L4 (mid-level), L5 (senior), or L6 (principal/staff). Every question, model answer, and scoring bar calibrates to that level. Aiming for Amazon NDE? Pick L5. NDEx or principal-track? Pick L6.
Answer in your own words. Use voice to catch the habits that hurt you in real interviews: the filler words, the hedges, the answers that trail off without conviction. Text is faster. Both produce a full scorecard.
A breakdown by category, a model answer, and flashcards built from what you missed — scheduled by spaced repetition so you see them again before you forget. The mastery heatmap tracks demonstrated depth across every topic.
FAQ
BGP path selection and policy, OSPF area types and adjacency, TCP/IP internals, DNS and DNSSEC, TLS handshake and certificate chains, load balancing, Linux networking, network automation with Python, and AWS networking at depth — Transit Gateway, Direct Connect, VPC routing, Route 53, ELB, and security groups vs NACLs. Advanced topics include VXLAN, EVPN, MPLS, SD-WAN, BFD, ECMP, STP, IPv6 migration, and troubleshooting methodology.
Yes. The behavioral module covers Amazon Leadership Principles with STAR-format coaching and LP-mapping feedback on every answer. Technical content covers AWS networking in depth — Transit Gateway design, Direct Connect redundancy, VPC routing edge cases, ELB listener rules, Route 53 routing policies. System design generates AWS-scoped architecture challenges. The JD Planner reads a real NDE or NDEx job post and builds a calibrated study plan from it.
Every answer is scored on five dimensions: technical accuracy, depth, structure, clarity, and conciseness. Voice answers add speech-specific metrics: filler words per minute, hedging phrases, pace, and trailing endings. You see the numbers, a model answer, and flashcards built from your specific gaps — not a pre-made deck.
Yes. Every drill, mock interview, and lab has a voice record option. The transcript is generated automatically and scored the same as text, with the addition of speech metrics: filler word count, words per minute, stutter count, hedging language, and a confidence signal score. Practising out loud surfaces delivery habits that writing never will.
Sharpen has three calibration levels: L4 (mid-level candidate), L5 (senior engineer), and L6 (principal/staff). At L4 the scoring bar rewards a correct, well-structured answer with one example. You do not need years of production experience to start. That said, Sharpen assumes you know what BGP and OSPF are and want to get better at explaining and applying them under pressure — it is not a first-exposure learning tool.
Anki asks you to recall a fact. Sharpen asks you to explain a concept to a bar-raiser, defend a design decision, or troubleshoot a live fault. These are performance tasks, not recall tasks. Sharpen does use spaced-repetition flashcards — but they are built from your own weak answers, not a pre-made deck, and they are one piece of a system that also includes probed mock interviews, branching labs, speech coaching, and a mastery heatmap.
You answer the opening question by voice or text. The AI interviewer then probes with 2–3 follow-ups based on what you actually said — if you skipped a trade-off, it asks about it; if you made a claim, it may probe the mechanism. At the end you get a full scorecard covering every turn. Depth and expected detail scale with your chosen level: L4, L5, or L6.
Yes. Create an account and get full access — no payment required.
Create an account. Pick your level. Answer one question. That is how it starts.
Start practising