We're Hiring

Build the missing layer between databases and agents.

Palantir charges $10M to connect the dots.

Yurii Rashkovskii

Founder, Inferal

There's a layer missing from the data stack. Most people don't notice because they've been working around it for so long it doesn't feel like a gap anymore. It feels normal.

Your database stores facts. Your application code reacts to requests. Between those two things, there's a question nobody answers well: what should happen right now, given everything the system knows?

Today, the answer is scattered across cron jobs, event handlers, polling loops, and message queues. Dozens of ad hoc mechanisms pretending to be one coherent system. They sort of work. Until they don't. And when they don't, it goes one of two ways: silence, where the thing that should have happened simply doesn't, or a cascading failure, where one missed condition tears through everything downstream.

What we're building

A system that watches everything your business knows and acts the moment conditions align. Not a workflow you define. Not a query you run. Continuous evaluation against streaming facts, where agents activate with full context already assembled. Every decision traceable. Every inference with provenance.

Think of it as the missing layer between your database and your agents.

The hard problems

These are real, open problems. We have approaches, not answers. If that distinction matters to you, keep reading.

Rule system engineering.: Building a rule system that people, AI agents, and automated pipelines can all author and reason about. Expressive enough to capture real business logic, structured enough to verify and optimize at scale.
Petabyte-scale data hypergraphs.: Facts and their relationships form hypergraphs that grow to petabyte scale. Incremental evaluation that doesn't recompute the world when one fact changes. Storage and traversal patterns that no off-the-shelf database handles well.
Real-time heterogeneous workload scheduling.: Rule evaluation, data ingestion, agent activation, and model inference all competing for resources with different latency requirements. Scheduling that adapts in real time without centralized coordination.

Where we are

Early stage. Paying customers, revenue, hard problems we haven't solved yet. The roadmap changes as we learn. You'll shape the product, not just build what's specced.

Who we're looking for

Founding Engineer

Someone who can hold database internals and agent orchestration in the same mental model. You've worked close to metal: graph engines, distributed scheduling, storage systems. You also know what it takes to make AI agents reliable in production, not just in demos. Rust, Python, and whatever else the problem demands. You're comfortable when the map is incomplete and the next step isn't obvious. That's where the interesting work lives.

Please note that at this moment, we're NOT looking for dedicated AI/ML engineers; using AI/ML is just a part of our routine.

We care more about what you've built than where you went to school. More about how you think through a problem than which languages are on your resume. Attitude first, then belief, then skill. Skills can be learned. Attitude can't.

Not ready to apply yet? Join the Software Internals Discord, a community where people dig into the kinds of problems listed above.

Ontologist (Contract or Full-time)

We're looking for an ontologist to help us make meaning explicit enough for software to act on. You'll design narrowly scoped ontologies for real operational contexts, then define the mappings that connect them to messy source systems. The work spans domain modeling, identity criteria, relationships, constraints, provenance, and safe translation between vocabularies.

We are not trying to produce one universal model or hand-build a bespoke artifact for every customer. We're building the processes and tooling to create, test, compose, and maintain a large library of scoped ontologies that can be reused across organizations and adapted to real operational contexts. You'll help turn ontology engineering from a consulting exercise into a repeatable product capability.

You can work with domain experts and engineers, distinguish a useful operational model from an elegant abstraction, and test definitions against counterexamples. You can write the scripts you need to inspect data, automate repetitive work, and validate or transform models, with or without LLM assistance. Experience with knowledge graphs, RDF, OWL, SHACL, semantic interoperability, or formal knowledge representation is useful, but judgment matters more than allegiance to a particular stack. This position is open on a contract or full-time basis.

What to expect

How we work is part of what we build. We keep finding better ways for people and agents to work together, then turn what works into shared tools and practice. Our work spans repositories, SaaS products, and systems we build ourselves. We use each where it works best, and you'll help shape that system.

We prioritize sleep. Great work comes from rested minds, not from heroic hours.

Remote-first, but we value face time. With coworkers, customers, peers. We make it a point to get together when we can.

Flexible engagement. The ontologist position is open to contract and full-time candidates. Full-time roles include a competitive salary and early options with extended exercise windows for tenured team members.