SDE Intern

About Inturview

Inturview is an AI interviewer designed for practicing interview rounds, including recruiter screens, behavioral stories, coding rounds, and system design. Companies utilize Inturview for first-round interviews, providing a consistent experience and scorecards for applicants. The company is early-stage with two founders, a working product, and an AI interviewer ready for launch.

Key Responsibilities

• Ship features end-to-end, including React/Next.js UI, API routes, Prisma schema changes and migrations, and tests.
• Work on the AI interviewer itself, focusing on prompts, conversation flow, follow-up questions, and progression.
• Harden the AI scoring pipeline, ensuring structured output from Claude that is validated, includes retries, and uses a durable Postgres job queue.
• Improve live interview experiences by addressing latency, interruptions, and reconnects for voice and video sessions using WebRTC, OpenAI Realtime, and LiveKit.
• Build evaluations to measure the impact of prompt changes on interviewer performance.

Requirements & Tech Stack

• Next.js
• TypeScript
• PostgreSQL
• Prisma
• Python / FastAPI
• Claude API
• OpenAI
• Realtime WebRTC / LiveKit
• Playwright
• GitHub Actions

Qualifications

• Currently pursuing a BS or MS in Computer Science or a related field, or equivalent shipped work.
• Available 15–20 hours a week for at least 3 months.
• Strong TypeScript and React.
• Experience building and shipping at least one full-stack web app.
• Familiarity with SQL and relational databases, preferably Postgres, including joins, indexes, and transactions.
• Proficient with Git and pull requests, including small commits, clear descriptions, and prompt review responses.
• Hands-on experience with an LLM API (Anthropic or OpenAI), including streaming, tool use, or structured JSON output, and failure handling.
• Experience with a daily AI coding workflow (e.g., Claude Code, Cursor, Copilot), including reviewing, testing, and correcting generated code.
• Experience with prompt design and providing evidence-based explanations for prompt effectiveness.

Nice to Have

• Next.js App Router
• Prisma
• Tailwind CSS
• Python and FastAPI
• WebRTC, audio streaming, or a realtime voice API
• An eval set or LLM-as-judge pipeline
• Playwright end-to-end tests

To Apply

Send a link to something you've shipped that uses an LLM, along with three to five sentences describing one issue encountered and its resolution. Email [email protected] with "SDE Intern" in the subject.

Internship Details

• Unpaid to start, with a path to paid roles upon achieving product-market fit or securing funding.
• Real ownership with work going to production, direct collaboration with founders, and the ability to showcase shipped work.
• Flexible schedule: Remote / City, 15–20 hours per week, starting [date]. Accommodations for classes and academic credit are available.