Want to know if this job is worth applying to?
New York, New York, United States
Onsite
Lirvana Labs builds Mission Control, an AI system for school leadership. School and district leaders today spend hundreds of hours a year pulling data out of disconnected systems by hand just to answer basic questions: is this intervention working, which students are slipping, what do we tell the board. Mission Control answers those questions in real time. Yeti Confetti Kids, our teaching and learning product, is an AI co-pilot for early numeracy and literacy, covering 200+ skills mapped to Common Core, CASEL, and Bloom’s, and giving educators a clear picture of each child’s cognitive, social, and emotional development.
Everything we ship is co-designed with the partners who will use it: NYC Public Schools, the Urban Assembly, TNTP, the Hong Kong Education Bureau, and the Joan Ganz Cooney Center at Sesame Workshop. That approach has earned recognition from MIT Solve, EdTech Week NYC, IEI (Superintendent’s Choice), and H&K EdCity, but the recognition that matters most is a principal, superintendents, and leaders telling us the product changed how they run their organization.
We started in 2021 and are backed by impact-focused investors: Kapor Capital, Transcend Capital Partners, and Chingona Ventures. We’re headquartered in Silicon Valley with global offices in New York, Singapore, and Hong Kong.
At the heart of our approach is our co-design process. Every conversation is an opportunity to build a long partnership focused on student impact and system wide change. Today, when a district tells us what they need, the path from that conversation to a working solution runs through several people. We want one person who owns it end-to-end.
You’ll sit in discovery with principals and school leaders, figure out what they actually need (often not what they first ask for), and then build it, which includes prototypes, agentic workflows, evaluations, and data pipelines on top of our platform. Then you carry it through to production and hand it off cleanly for our platform team to maintain. A working prototype closes deals here and builds trust with partners, so you’ll co-own the pipeline with Sales.
Roughly how your time breaks down:
Half building. Prototypes, POCs, custom agentic tools and eval scripts. You own them from design through deployment and handoff.
A quarter in discovery. Running customer conversations alongside Sales and earning technical credibility and trust with partners with your knowledge around our solutions and architecture
A quarter of requirements. Turning what you heard into something shippable, scoping custom workflows, and calling what fits on our platform and what doesn’t and working with our platform and product team on roadmap dependencies.
You report to the Head of Product and work day to day with Sales, Engineering, and DevOps.
What this looks like in practice
A few examples of the kind of work you’d own:
A district wants to know which 9th graders are at risk of not graduating, but their attendance, grades, and behavior data live in three systems that have never talked to each other. You design the pipeline, build the early-warning logic on Mission Control, and sit with their data lead until it’s running on their real data.
A school network wants an agent that drafts comprehensive education plans from a student’s recent data. You scope it with the principal, build a working version, run evals against what their team would have written, and iterate until it’s good enough to put in front of staff.
How we work
Small team, high ownership, biased toward action. We co-design and build with partners, show them unfinished prototypes, and change course when we learn something. Nobody will manage you closely; you’ll be trusted to make calls and expected to own the outcomes.
The mission is the reason most of us are here. The people on the other end of this work are school leaders responsible for thousands of students, and the students themselves, many of them multilingual learners in under-resourced schools. The work is hard and the systems are old and the data is a mess, and when it works, a kid who was going to fall through the cracks doesn’t. We take that seriously.
Why this role, specifically
You’ll own the full arc from first conversation to production, which almost no role at a larger company lets you do.
You’ll build agentic systems on real data for real users, not demos.
Your work will directly shape the roadmap. When you find a gap, you’re in the room deciding whether we fill it.
You’ll get paid on the deals you help close, on top of base and equity.
4+ years shipping production software, including time as the technical person in the room with customers.
You can take a user story from a non-technical stakeholder, turn it into something buildable, and build it yourself. You’re comfortable owning code in production.
Snowflake and modern data engineering: pipelines, ETL, data modeling, SQL optimization. School data is messy and you’ve worked with messy data before.
You’ve built n8n workflows and know how agentic workflows plug into a data stack.
You’ve sold, or been close to the sale: multiple stakeholders, objection handling, keeping a deal moving with leadership on the other side.
You can run a project end to end and keep the timeline honest.
You make day-to-day technical calls without asking permission, and you have the judgment to make them well.
You explain technical things plainly. Superintendents and principals are time strapped leaders and you need to get to the point.
You’ve led a small team or mentored junior full-stack engineers.
You’d rather ship a scoped version than polish a slide.
Nice to have: forward-deployed engineering experience, or you’ve started something or been early at something.
This exercise is very important and is the first step in the evaluation process.
