The setting
makes software for building and testing intelligent vehicles and autonomous machines. turns real-world driving data into photorealistic simulation environments for training and validating autonomous driving systems.
Lane-change evidence looks cleaner when the cut-in cases disappear.
Fast evaluation results protect iteration speed, while retained cut-in coverage protects the meaning of a planning regression.
Who you’d be doing this for
“I found the cut-in drive in the source set, then spent an hour proving it never reached the run.”
Jihoon Kobayashi · Autonomous Vehicle Validation Engineer · a character in this brief
Runs closed-loop lane-change regression campaigns for an automotive OEM using reconstructed and synthetic driving scenarios.
What is at stake
Approved cut-in records are missing from lane-change evaluation runs, making regressions appear smaller than they are. You must weigh rapid metric recovery against admitting records whose automated labels have not been verified.
Why it isn’t already fixed
Every obvious fix costs something else. That’s the part you’d have to decide.
- evaluation speed vs. coverage integrity
- automated labels vs. verified inclusion
- planning iteration vs. regression credibility
- compact reporting vs. traceable exclusions
Why Applied Intuition
Neural Simulation turns real and synthetic driving data into the evidence OEM validation engineers use to judge closed-loop autonomy behavior.
Written with these in mind
Not your kind of problem? 93 more at Applied Intuition, or browse every organization.
Where the organization context comes fromThe postings this brief was written from, and the reporting read before it.
The organization context comes from public information. The situation and characters are written for the brief.
Job postings it was written from
Public descriptions of the work at Applied Intuition. They are listed as sources, not as openings. The lines under a posting are its own words, where it meets this brief.
- Software Engineer (SDS Core - Evaluation) (opens in a new tab)
- Build and curate the datasets — real and synthetic — that those metrics run against
- You'll work closely with stack developers to turn their open questions into rigorous metrics, drawing on both real road data and novel synthetic scenario generation to evaluate behaviors too rare or too dangerous to encounter on the road.
- Software Engineer - Data Pipeline (opens in a new tab)
- Mapping Specialist · no longer posted
- A strong understanding of lane-level geometry at CM precision, routing, and HD (High-Definition) map structures used in autonomous driving.
- Senior Software Engineer - Data Engine (opens in a new tab)
- As a data engineer experienced in machine learning, you will be responsible for both our edge application which runs perception stacks and our cloud data engine which ingests video data to run the next iteration of the model before we deploy it back down to the edge.
- Engineering Manager - Autonomy Storage Infra (opens in a new tab)
- Drive engineering decisions around scalable data ingestion, indexing, storage, and retrieval for high-volume autonomous vehicle and physical AI workloads
This is the setup. The work is inside.
Running it puts you in the room: the full situation and its constraints, stakeholders who push back in their own words, and the decisions that are yours to make. What you produce becomes a Day One Plan — work you can show someone instead of describing.
Where similar work appears
Other live briefs connected to this one, and what connects them.
Similar problems at other organizations
Like this one, each of these turns on fixing one thing without breaking another.
Wet-route boundary inconsistency rose from 1.1% to 8.4% after the latest model update.
You must weigh a fast model-side correction against preserving the aggregate gain.
Duplicate retries are making rare behavior coverage appear larger than it is.
You must weigh fast stream recovery against evidence reviewers can trust.
Dock-site searches are omitting captured flights from the curated dataset.
You must weigh a quick schema mapping against preserving source-level provenance.
The same setting, from another role
Each is set around Applied Intuition’s Neural Simulation, written for a different role. They’re separate briefs, not the same scenario.
A known Kubernetes deployment failure keeps pushing scheduled simulation regressions past their window.
You must choose what AI can classify and run, and where an engineer must take over.
Validation engineers are rebuilding cloud setup details before comparable closed-loop campaigns can begin.
You have to weigh fast self-service against the controls that make reruns defensible.
A 12,000-run validation campaign is about to lose its AWS GPU capacity, and paired reruns are not fully matching.
You have to weigh a protected review window against a cutover that could alter trusted evidence.
Other problems with a similar purpose
In each of these, as in this one, the aim is to engineer security-critical controls whose correctness holds up even when carefully examined by skeptics.
Held content can be classified in most samples, but nothing proves that every retrieval surface enforces the split.
You have to ship enforcement, define the proof, and get sign-off before the irreversible migration.
A controller trust-path concern must be addressed within seven days across occupied sites.
You must weigh stronger remote protection against the chance of disrupting reliable door access.
An automated grader for AI-generated Python passes 97% of tests, but reviewers found real errors in code it marked safe.
You have to decide what evidence clears this for launch, since the code runs against customer warehouse data.
