Solution Engineer

You will be embedded inside our customers' hardest problems, and you will own the outcome until it works in production. You'll sit with executives, domain experts, and data teams to find the decisions that actually move their business - inventory that's in the wrong place, risk concentrations nobody can see, fraud patterns that only emerge across three systems, capacity plans built on guesses. Then you'll model their world in our ontology, formulate the reasoning problem, write the PyRel, and ship something that runs against their real data in their own Snowflake account. Every engagement here produces two deliverables. The first is the one the customer sees: a working decision system that changes how they operate. The second is the one that matters most to us: the pile of things you had to invent because our platform didn't have them yet. The modelling pattern you hand-rolled. The constraint formulation that should have been a primitive. The three-hour workaround where an API should have existed. You bring those back, you argue for them, and the strongest of them become product. That second deliverable is why this role exists. If you only ever deliver the first one, we've hired a consultant. We're not hiring consultants. You'll operate with unusual autonomy: you decide what's worth building, when a workaround is acceptable and when it's technical debt we'll regret, and when to tell a customer their real problem is not the one they asked about. You'll be technical enough that when something breaks in a customer environment, you find the root cause yourself rather than filing a ticket and waiting. • Own outcomes end to end - discovery, modelling, implementation, performance tuning, production hardening, and the measurement that proves it worked. Not a handoff at each stage. Yours. • Build, not describe - design and ship decision solutions on our modelling, reasoning, and learning stack: ontologies over customer data, rules, graph analytics, optimisation formulations, predictive models • Fill the gaps yourself - when a customer workflow is blocked on something the platform doesn't do, scope it and build it. Then push the general version upstream: read our source, form a hypothesis before you escalate, open the PR. • Close the loop with Product - every deployment generates a signal. Bring back reproductions, patterns, and specific failure modes ("the only way I could express this was by abusing X in this way"), not vibes. You are one of the loudest inputs int

You thrive in ambiguity and move with intent. You're motivated by deep understanding and meaningful impact. • Owner, not participant. You take full accountability for the outcome, not your slice of it. When something is broken and it's nobody's job, it becomes yours. • You build. Your instinct in the face of a hard problem is to open an editor, not a deck. You'd rather show a working prototype on real data than a diagram of one. • High conviction, low ego. You argue hard for what you believe, you're direct about what's wrong, and you change your mind quickly when the evidence turns. You challenge ideas without making it personal - people leave arguments with you feeling sharper, not smaller. • Rigorous. You root-cause things. You can explain both why it broke and why your fix works. Surface symptoms don't satisfy you. • Fast in unfamiliar territory. Dropped into a codebase, a domain, or a data model you've never seen, you're useful within days. "I only do backend" and "that's not my job" are phrases you don't use. • High tolerance for friction. Enterprise environments are messy - broken data, VDI access, security reviews, politics. You route around it and keep shipping. • Impact-driven. You want the thing you built to still be running, and still be load-bearing, two years from now. • 5+ years building and shipping production software, at least some of it inside customer or partner environments • Demonstrated end-to-end ownership: you have personally taken something from an ambiguous problem statement to running in production, and you can walk us through the whole arc, including what went wrong • Strong SQL and deep familiarity with cloud data platforms (Snowflake, BigQuery, Databricks, Redshift) • Strong programming ability - Python primarily; comfort with declarative or logic-style languages is a real advantage • Comfortable reading unfamiliar source code, interpreting stack traces, and debugging systems you didn't write • Able to hold your own with both a VP of Supply Chain and a staff data engineer, in the same meeting • Comfortable operating in high-autonomy, high-velocity, low-instruction environments • Built analytical, decision, or reasoning applications that reached production and stayed there • Experience with optimization, constraint solving, rule engines, graph algorithms, or ML on structured data • Semantic modelling, data pipelines, and governance in real enterprise settings • Track record of upstream contribution - features, tools, or abstrac

Who We Are At RelationalAI, we’re solving one of the most important challenges in artificial intelligence: how to teach large language models the logic, semantics, and business context of the modern enterprise. Frontier models are trained almost entirely on public data — they can speak about the world, but they don’t understand your business. We fix that. RelationalAI has pioneered a breakthrough called Superalignment — technology that enables LLMs to learn natively from private, structured enterprise data inside the data cloud. By combining this with relational knowledge graphs and our proprietary neuro/symbolic-relational reasoners, we deliver trustworthy decision intelligence: systems that use semantic models to truly understand how a business operates and can reason across its data to drive better outcomes. We’re a globally distributed team of engineers, scientists, and builders redefining how AI learns from data. We believe that high-stakes decisions deserve frontier intelligence — intelligence that’s explainable, aligned, and grounded in reality. If you’re driven by curiosity, thrive in complexity, and want to help build the system that brings true understanding to enterprise AI, you’ll feel right at home here. The Role You will be embedded inside our customers' hardest problems, and you will own the outcome until it works in production. You'll sit with executives, domain experts, and data teams to find the decisions that actually move their business - inventory that's in the wrong place, risk concentrations nobody can see, fraud patterns that only emerge across three systems, capacity plans built on guesses. Then you'll model their world in our ontology, formulate the reasoning problem, write the PyRel, and ship something that runs against their real data in their own Snowflake account. Every engagement here produces two deliverables. The first is the one the customer sees: a working decision system that changes how they operate. The second is the one that matters most to us: the pile of things you had to invent because our platform didn't have them yet. The modelling pattern you hand-rolled. The constraint formulation that should have been a primitive. The three-hour workaround where an API should have existed. You bring those back, you argue for them, and the strongest of them become product. That second deliverable is why this role exists. If you only ever deliver the first one, we've hired a consultant. We're not hiring consultants. You'll operate with unusual autonomy: you decide what's worth building, when a workaround is acceptable and when it's technical debt we'll regret, and when to tell a customer their real problem is not the one they asked about. You'll be technical enough that when something breaks in a customer environment, you find the root cause yourself rather than filing a ticket and waiting. What You'll Do • Own outcomes end to end - discovery, modelling, implementation, performance tuning, production hardening, and the measurement that proves it worked. Not a handoff at each stage. Yours. • Build, not describe - design and ship decision solutions on our modelling, reasoning, and learning stack: ontologies over customer data, rules, graph analytics, optimisation formulations, predictive models • Fill the gaps yourself - when a customer workflow is blocked on something the platform doesn't do, scope it and build it. Then push the general version upstream: read our source, form a hypothesis before you escalate, open the PR. • Close the loop with Product - every deployment generates a signal. Bring back reproductions, patterns, and specific failure modes ("the only way I could express this was by abusing X in this way"), not vibes. You are one of the loudest inputs into our roadmap. • Run technical discovery that gets to the truth - workshops, demos, and proofs of concept designed to find out whether we can actually solve the problem, not to look impressive. • Leave things better than you found them - document as you go, in the repo, same week. Turn one-off work into reusable reference implementations so the next person starts where you finished. No branch of yours should be diverging for a month. • Refuse shortcuts that compound - no undocumented config drift, no "it works now" without knowing why it broke, no restarting the service before you've captured the evidence. Who You Are You thrive in ambiguity and move with intent. You're motivated by deep understanding and meaningful impact. • Owner, not participant. You take full accountability for the outcome, not your slice of it. When something is broken and it's nobody's job, it becomes yours. • You build. Your instinct in the face of a hard problem is to open an editor, not a deck. You'd rather show a working prototype on real data than a diagram of one. • High conviction, low ego. You argue hard for what you believe, you're direct about what's wrong, and you change your mind quickly when the evidence turns. You challenge ideas without making it personal - people leave arguments with you feeling sharper, not smaller. • Rigorous. You root-cause things. You can explain both why it broke and why your fix works. Surface symptoms don't satisfy you. • Fast in unfamiliar territory. Dropped into a codebase, a domain, or a data model you've never seen, you're useful within days. "I only do backend" and "that's not my job" are phrases you don't use. • High tolerance for friction. Enterprise environments are messy - broken data, VDI access, security reviews, politics. You route around it and keep shipping. • Impact-driven. You want the thing you built to still be running, and still be load-bearing, two years from now. What This Role Is Not We'd rather be blunt than waste your time: • It is not demo-and-handoff pre-sales. You don't disappear after the POC; you're there when it goes to production. • It is not staff augmentation. You own outcomes, not hours or ticket queues. • It is not adv