HN Showcase
Back to browse
TechGrill – AI Interview Practice

TechGrill – AI Interview Practice

by wasivis·Feb 26, 2026·1 point·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidSolve My ProblemShip It

Sub-second LLM latency keeps interview pressure realistic; junior-to-senior difficulty scaling.

Strengths
  • •Groq's 70B Llama inference enables real-time conversational flow without awkward pauses
  • •Difficulty tiering (Junior/Mid/Senior) with performance scorecards tailored to level
  • •Transcript persistence lets users review weak points and track improvement
Weaknesses
  • •Highly competitive space: LeetCode, AlgoExpert, Pramp, and dozens of AI interview coaches exist
  • •No differentiation shown—unclear what makes TechGrill's questions or scoring better than alternatives
Category
Education
Target Audience

Software engineers preparing for senior-level technical interviews

Similar To

Pramp · AlgoExpert · LeetCode Premium

Post Description

I wanted to simulate the pressure of a real senior-level, cold, brief, and focused on edge cases interview, so I built TechGrill.

The Stack:

- Next.js 15 & Server Actions

- Groq (Llama 3.3 70B): Used for sub-second latency to keep the "grilling" conversational.

-Neon (Postgres) & Drizzle: For transcript persistence.

It handles Junior to Senior levels and generates a performance scorecard after a few questions.

Any feedback is appreciated!

Similar Projects