AI Safety & Alignment Engineer
(Systems)
Location
Work-from-Anywhere Policy
Employment Type
Full-Time
Department
Engineering
Base Compensation
$260K - $340K USD +
Offers Equity + Acceleration Options. Performance Bonus up to 50% of base.
Total Compensation
$320K - $540K USD +
Offers Equity + Acceleration Options.
Additionally, this role is eligible to participate in our accellerated equity plan, culture and benefits program. Equity calculators and financial education programs are included with these programs.
Benefits include, but not limited to: Comprehensive health, dental, vision and mental health coverage, retirement benefits (401K match up to 10%), and flexible PTO, together with fully funded course programs.
Why Instant.Lawyer
Instant.Lawyer is a hypersecure platform at the intersection of AI and blockchain, built to augment productivity, automate complex workflows, and improve how decisions are made in the real world. We’re part of a growing ecosystem including Instant.Tax and Instant.Accountants, leveraging reasoning-capable LLMs customized by our team of lawyers, accountants, engineers, and scientists.
Our software combines AI, machine learning, cryptography, and incentive systems into a single architecture designed to operate continuously and reliably. It is researched, built, and tested by a team that understands both the technical and real-world consequences of decision-making systems. With strong early traction and growing demand, we are scaling deliberately and hiring people who want to build infrastructure that matters.
Here's why Instant.Lawyer needs you:
​
​We are working on a simple idea with significant implications. Technology should improve life for people, families, businesses, and communities. Not occasionally, but continuously. That standard starts with the team. We invest in our people so they can build systems the world can depend on.
​
-
We are building safe, interoperable, decentralized AI systems designed to operate across domains, environments, and real-world constraints;
-
We have strong product-market fit and are working with leading organisations across AI, finance, and legal services;
-
We are backed by global investors who understand the scale of what is being built and support long-term execution;
-
World-Class Team: We’re hiring the best talent from Anthropic, OpenAI, Google, Stripe, Tesla, and more, blending expertise across engineering, law, and design.
-
Partnerships:Our engineers and researchers collaborate directly with leading model providers including OpenAI, Anthropic, Google Gemini, and Meta LLaMA.
-
We offer packages that are "North of Fair" to attract and retain world-class talent to work alongside each other.
Role Overview
As an AI Safety & Alignment Engineer, you will define how correctness is enforced across the Instant.Lawyer system.
​
We are building systems that do not just generate outputs, but influence decisions and take actions across legal, financial, and real-world domains. In this environment, capability is not enough. The system must behave correctly under uncertainty, constraint, and scale.
​
You will design and implement the mechanisms that make this possible.
​
This includes translating principles such as acting in a participant’s best interests into executable constraints, evaluation systems, and continuous verification.
​
You will work at the boundary between model behaviour and system behaviour. The focus is not on individual responses, but on ensuring the system operates reliably over time, across changing conditions and edge cases.
​
This role is central to how Instant.Lawyer is built. Safety is not added after the fact. It is embedded directly into the architecture.
What You'll Do
-
Design and implement evaluation frameworks that measure correctness, reliability, and failure modes in production
-
Build constraint and policy systems that govern how models behave in real-world scenarios
-
Develop methods to detect and reduce: hallucinations, unsafe or misleading outputs, bias and inconsistency.
-
Define and operationalise alignment metrics beyond accuracy, including decision quality and outcome consistency
-
Integrate alignment directly into LLM pipelines (RAG, tool use, multi-step reasoning)
-
Build monitoring and feedback systems to track behaviour continuously, not episodically
-
Investigate edge cases and failure scenarios, and implement system-level corrections
-
Contribute to systems that are auditable, explainable, and verifiable
What You Have
-
4+ years of experience in applied machine learning, AI systems, or AI safety, with a track record of shipping production systems.
-
Hands-on experience building, deploying, and maintaining LLM-based systems in production, not just experimentation or research
-
Experience designing or implementing: evaluation frameworks, guardrails or policy systems or constraint-based decision logic.
-
Strong understanding of: LLM behaviour, hallucination risks, bias, drift, and failure modes in real-world environments.
-
Experience working with multi-step LLM systems, including: RAG pipelines, tool use or agent-based workflows.
-
Proficiency in Python and experience with modern ML frameworks (e.g. PyTorch, JAX, TensorFlow).
-
Experience defining and measuring system performance beyond accuracy, including: reliability, consistency, and behaviour under edge cases.
-
Ability to translate high-level requirements into testable, enforceable system behaviour.
-
Experience working cross-functionally with engineers and domain experts to ship systems into production
Compensation Range
$250,000 - $300,000 USD
Instant.Lawyer is an equal opportunity employer and does not discriminate on the basis of race, gender, sexual orientation, gender identity/expression, national origin, disability, age, genetic information, veteran status, marital status, pregnancy or related condition, or any other basis protected by law.