Careers

We hire opportunistically for full-stack, AI/ML, and design. One open role today — see below. For everything else, send a note to careers@flowshark.ai.

AI Intern — Applied LLM Engineering

Type: Internship (full-time or part-time, 3–6 months)  ·  Location: Remote  ·  Compensation: Stipend + conversion path to full-time

About the role

Work directly on the AI surfaces that drive Flowshark — proposal generation, lead scoring, Finn (RAG-grounded assistant), template inheritance, and the AI-zone regenerate pipeline. You will own end-to-end features: prompt design, retrieval pipeline, worker job, cost accounting, and the React surface that exposes it.

What you will work on

• AI zones inside templates — improving the ProseMirror ↔ DOCX bridge so generated sections render with bold, tables, and lists intact across PDF + DOCX export.
• Lead scoring + auto-routing — feature engineering and prompt tuning on inbound email threads. Background jobs run through PGMQ workers.
• Finn assistant — RAG over tenant CRM + email + knowledge-base, served via SSE streaming. Retrieval quality, citation grounding, refusal behavior.
• Cost & credit accounting — every AI action charges a credit via PricingService + WalletService. Help us keep the worker → hold → finalize flow correct as new actions ship.

Tech stack you will touch

• Backend: Python 3.12, FastAPI, async SQLAlchemy, Alembic, Pydantic v2.
• Data: Postgres (Supabase-managed), row-level security, PGMQ for background queues.
• AI: Anthropic + OpenAI APIs (no-training tier), structured-output prompting, vector retrieval, SSE streaming.
• Document pipeline: python-docx, weasyprint, cairosvg, custom ProseMirror ↔ DOCX converter.
• Frontend: Next.js 14 App Router, TypeScript, React Query, TipTap editor, Tailwind, next-intl.
• Infra: Render (app hosting), Supabase (Postgres + auth), Stripe (billing).

What we look for

• You have shipped something with an LLM — a side project, a research demo, a hackathon win. We care about what you built more than where you studied.
• Comfortable in Python and TypeScript — you will cross the stack daily.
• Strong instinct for prompt design and evals. You can articulate why a prompt works, not just that it does.
• Bias toward shipping. We merge to main several times a day; interns ship to production in week one.
• Bonus: experience with RAG, structured output (JSON-mode / tool use), or document parsing (DOCX, PDF, OCR).

What you get

• Direct ownership of features that real customers use.
• Mentorship from founders working hands-on across backend, AI, and product.
• Monthly stipend + Flowshark credits to ship side projects.
• Strong-performance interns get a full-time offer at the end of the internship.

How to apply

Email careers@flowshark.ai with subject AI Intern — <your name> and include:

• A short note (3–5 sentences) on the most interesting LLM-driven thing you have built.
• Links — GitHub, a deployed demo, or a write-up. Code beats résumés.
• One paragraph on which area above (AI zones / lead scoring / Finn / cost accounting) you want to own first, and why.

We read every application and respond within 7 days.