Founding Full Stack Engineer

Founding Full Stack Engineer — AI Infrastructure for Trial AttorneysRemote US · Full-time$180K–$220K base + 0.20%–0.50% equityThe company

This company is building a high-performance agentic litigation engine for trial attorneys.

The product combines 20+ million embedded litigation documents, proprietary judicial behavioral intelligence, and a coordinated swarm of specialized AI agents that research, draft, cite-check, and simulate the judge in a single end-to-end workflow.

This is not a generic chatbot layered onto legal workflows. The system has to produce output that is grounded, auditable, and defensible when the work is scrutinized by experienced litigators.

The team is still small: 3 full-stack engineers and 2 data engineers, with direct access to the founder and enough traction to justify building for real production use, not a prototype path.

Recent execution has been concrete: the admin billing surface was scoped and shipped in under a week.

The role

This is a founding engineer seat for someone who wants broad ownership across the primary product surface and the AI system underneath it.

You will be joining as founding engineer #3, and you will work directly with the founder and the small engineering/data team.

Your scope spans frontend craft, full-stack feature delivery, third-party integrations, and agent architecture. Some weeks will be UI-heavy. Some weeks will be deep in orchestration, evals, and failure handling. Most weeks will include both.

This role is about owning the experience attorneys see, the systems that generate the output, and the reliability constraints that decide whether the product is trusted in practice.

The technical problem

Trial attorneys do not tolerate vague answers, hidden source chains, or brittle workflows.

The product has to ingest and normalize legal data from external systems, route work through specialized agents, extract and structure documents, keep retrieval grounded, and expose results in a UI that makes review and correction straightforward.

The hard part is not wiring up an LLM API.

The hard part is building orchestration, evals, source boundaries, and review UX that make AI output auditable, repeatable, and accurate enough for high-stakes litigation work.

What you'll own

• The primary AI interface: build the frontend lawyers actually use in React, Next.js, and TypeScript, with a focus on clarity, responsiveness, and reviewability.
• Review UX: design the workflows that let attorneys inspect sources, verify citations, correct output, and move from answer to action without losing context.
• Full-stack feature delivery: ship product work across frontend, backend, and AI systems without waiting on handoffs.
• Agent architecture: design orchestrators, sub-agents, tool-use flows, routing logic, and failure handling for multi-step legal workflows.
• Evaluation and reliability loops: build the test harnesses, regression checks, and output review flows that catch bad citations, weak grounding, and brittle tool behavior before customers do.
• Document extraction and RAG: develop the pipelines that turn messy legal inputs into structured, retrievable, source-linked data.
• Third-party integrations: own integrations with Clio, iManage, Salesforce, and other systems end to end, from scoping through deployment and edge-case handling.
• Backend quality and performance: improve the stability, observability, and upgrade path of the system as usage grows.
• Small-team leverage: work directly with the founder and the engineering/data team to turn ambiguous product goals into shipped systems.

Who this is for

Nominally, this role is scoped for engineers with 3–8 years of production experience. In practice, the bar is senior-level ownership of systems that already matter to users.

You are likely a strong fit if you have:

• Shipped production software across frontend and backend, not just one side of the stack.
• Owned a product surface or technical system end to end, with real accountability for quality, not just task completion.
• Strong judgment on UX craft, especially where interface design affects user trust and workflow speed.
• Experience with agentic systems, tool use, retrieval, or evaluation loops, or the ability to ramp quickly into that space.
• Comfort debugging the full path from user action to model output to tool call to database state.
• Enough backend maturity to reason about reliability, data modeling, and integration failure modes.
• Strong opinions about correctness, traceability, and how to keep AI systems inside source boundaries.
• The ability to work with incomplete requirements and still make practical architecture decisions.
• The habit of raising the bar on product quality without needing a heavy process layer.
• You already use AI coding tools to accelerate your own output, but you still review generated code like production code.

Tech stack

• Frontend: React, Next.js, TypeScript
• Product surface: a primary AI interface plus supporting workflows and integrations
• AI layer: agent orchestration, sub-agents, tool use, RAG, extraction, evals
• Integrations: Clio, iManage, Salesforce, and other third-party systems

The backend and system layer are part of the role, but the hiring signal is not narrow specialization. The expectation is that you can move across the product surface and the AI system with enough depth to make the right tradeoff, then ship it cleanly.

Why now

The company is pre-Series A and already has enough signal to work on the real problem: making AI dependable inside a legal workflow that has to survive scrutiny.

The next phase is not more experiments. It is building the product and infrastructure layer that can support scale: stronger integrations, cleaner orchestration, better evals, and a frontend that attorneys trust under pressure.

The team has already shown it can ship with speed. The harder opportunity now is to turn that speed into a durable system that stays correct as scope expands.

This role is not for you if

• You want a narrow, pre-scoped ticket queue.
• You prefer to stay on one side of the stack.
• You are uncomfortable owning UX quality in a product where trust is part of the technical problem.
• You want AI work that stops at prompt iteration and demo quality.
• You treat reliability, citations, and integration edge cases as someone else’s problem.

Compensation and logistics

• Base salary: $180K–$220K
• Equity: 0.20%–0.50%
• Employment: full-time
• Workplace: Remote US-based, with a strong preference for New York City
• Timezone: Eastern Time overlap preferred, at least 4 hours daily
• Visa sponsorship: not available; H1B transfers and TNs only
• Vest: 4 years with a 1-year cliff

About AuroraAurora helps exceptional engineers find the right role at some of the most ambitious startups worldwide.

We work with teams that hire for ownership, technical depth, and clear scope.