Mid-Level Full-Stack Engineer
Architect and ship production systems for clients across AU, US, EU, and the Middle East — Next.js platforms, AI-native data layers, and the deploy pipelines that keep them running.
- Location
- Karachi · Remote-friendly
- Type
- Full-time
- Compensation
- Above-market PKR base · per-engagement completion bonus · annual review
You’ll own two to three concurrent client engagements at any time, scoped between $10K and $50K. The work spans data modelling, application code, infrastructure, and the LLM integration layer — all of it shipped to production, monitored, and supported by you. We don’t train models. We do design retrieval systems, agentic workflows, and the database structure that makes them work. You’ll partner with the founder on architecture from week one — no PM layer between you and the client, no junior tier behind you to clean up after.
- Next.js
- TypeScript
- Postgres
- pgvector
- Tailwind
- GCP / AWS
- Claude API
What you’d own
- Own end-to-end delivery on two to three concurrent engagements — schema design, application code, infrastructure, deploy, and first-30-day incident response.
- Design retrieval pipelines, eval harnesses, and observability for LLM-integrated features running in production.
- Lead architecture conversations with client CTOs and founders; document trade-offs and decisions in writing.
- Hold the line on production quality — performance budgets, accessibility, security review, and eval coverage are non-negotiable.
- Partner with the founder on engineering hiring and code-review standards as the team grows past three.
What we’re looking for
- 3+ years shipping TypeScript in production, with at least one system you can walk an interviewer through from data model to deploy.
- Deep Next.js App Router experience — server components, streaming, edge runtime, and the caching trade-offs that come with them.
- Strong Postgres fundamentals — indexing, query planning, schema migration discipline, and an opinion on when to reach for pgvector.
- Production exposure to at least one LLM API (Claude, OpenAI, or equivalent) — prompt design, retrieval, eval frameworks.
- Owned a deploy pipeline on Vercel, GCP, or AWS — CI/CD, environment management, and the on-call rotation that sits behind it.
- Clear technical writing. You can put an architecture decision in a doc that a non-engineering stakeholder can follow.
Nice to have
- Open-source contributions, conference talks, or a maintained side project that demonstrates twelve months of sustained engineering attention.
- Production experience with pgvector, Pinecone, or another vector store at non-trivial scale.
- Hands-on with Vapi, Twilio, LiveKit, or another real-time / telephony-grade system.
- Prior agency or consultancy experience — comfort scoping, estimating, and pushing back on creep.
Apply for Mid-Level Full-Stack Engineer.
One form. Real human reads it. Three business days to a real response.
No agencies
We hire direct. No recruiters, no contingency placements, no referral fees.
Not the right seat?
See what else is open.