Work on AI systems for multifamily demand, conversion, and intelligence
Early-stage team. Hard systems problems. Real operational consequences. We are building AI-native infrastructure for multifamily operators.
AI infrastructure for the full leasing demand cycle
Renter discovery is fragmenting across AI search, social platforms, and distributed digital channels. Most leasing software was built for workflow management after the lead arrives — not for this environment.
We are building the infrastructure layer that connects demand capture, conversion automation, and demand intelligence. This means real AI systems, real signal pipelines, and real operator workflows — not demo-only product surfaces.
The systems we are building
Content generation pipelines
AI systems that create structured, discoverable property content for AI search and social channels
Signal capture infrastructure
Pipelines that process renter behavioral signals and score conversion probability in real time
Response orchestration
Event-driven systems that connect inbound demand to automated response and scheduling workflows
Demand intelligence models
Ranking and prioritization logic that helps operators focus on highest-intent prospects
PMS integrations
Structured integrations with Yardi, RealPage, Entrata, and other operator systems
Current openings
Senior AI/ML Engineer
Work on content generation systems, structured property data workflows, ranking and prioritization logic, and intelligence models built around renter demand signals.
What you will work on
- Build and improve AI content generation pipelines for structured property data
- Design ranking and prioritization models based on renter intent signals
- Develop demand intelligence systems that score prospect conversion probability
- Work on structured data workflows that improve property visibility in AI search environments
→Your work directly affects how operators discover and prioritize their highest-value prospects.
Backend Infrastructure Engineer
Build APIs, orchestration layers, signal pipelines, event flows, and system integrations that connect inbound demand to operator workflows.
What you will work on
- Design and build signal capture pipelines that process renter behavioral data at scale
- Build orchestration layers that connect demand channels to operator response workflows
- Develop API infrastructure that integrates with PMS systems (Yardi, RealPage, Entrata)
- Build event-driven architectures that enable real-time prospect engagement
→Your systems are the connective tissue between renter demand and operator action.
Full-Stack Product Engineer
Build operator-facing interfaces, internal tools, and workflow surfaces for response automation, scheduling, and demand visibility.
What you will work on
- Build operator dashboards that surface demand signals and prospect prioritization
- Develop scheduling and response workflow interfaces that leasing teams actually use
- Create internal tooling for content generation, review, and deployment
- Build data visualization surfaces that translate leasing signals into operator decisions
→Your interfaces are the primary way operators interact with the Valis platform.
Don't see a role that fits? Send us your background. We review all applications from engineers interested in AI infrastructure and real estate operations.
What to expect
Stage
Early-stage / product-building
We are building core infrastructure, not optimizing an existing product.
Work style
Remote-first
US timezone preferred. Async-first communication with regular team syncs.
Compensation
Salary + equity
Competitive salary with meaningful equity for early team members.
Ownership
High
Early team members own significant parts of the system with direct product influence.
Who thrives here
- Engineers who want to own systems end-to-end, not just implement tickets
- People who are comfortable with ambiguity and can move fast without losing quality
- Builders who care about the operational reality of what they are building
- Engineers who want to work on AI systems with real-world consequences, not toy problems
- People who prefer a small, high-trust team over a large engineering org
Who is not a fit
- Engineers who need a large team and defined processes to be productive
- People looking for a stable, low-risk environment
- Candidates who want to work on AI as a feature, not as core infrastructure
- Engineers who are not interested in understanding the operational problem they are solving
What I am actually looking for
"I spent 16 years in hospitality and digital media before building Valis. I know what operators actually deal with — and I know how far most PropTech products are from solving the real problem.
I am not looking for engineers who want to work on AI in the abstract. I am looking for people who want to build real systems that work in production environments, with real leasing teams, at properties where the stakes are actual.
We are based in Chapel Hill, NC. We are early-stage, AI-native, and building from 0 to 1. If that sounds like the right environment for you, let's talk."
— Jinbo Chen, Founder & Director
What we are building with
Frontend
React / Next.js · TypeScript · Tailwind CSS
Backend
Python · FastAPI · Node.js
Database & Auth
Supabase (PostgreSQL + RLS) · Row-level multi-tenant isolation
AI & Orchestration
Cursor (AI-native dev) · Vellum AI / n8n for agent routing · OpenAI / BytePlus Seedance
Compliance Layer
Zillow BERT classifier for FHA/ADA guardrails
Attribution
TikTok UID ↔ Supabase Guest Card stitching algorithm
Integrations
Yardi · RealPage · Entrata (PMS API layer)
Deployment
Vercel · Supabase Cloud · GitHub Actions
We are an AI-native team. 80%+ of our development workflow runs through Cursor (Composer/Plan Mode). We expect engineers to have strong prompt engineering skills and the TypeScript/Python depth to catch and fix what AI gets wrong.
Why this work is worth doing
Real category-building work
You are not joining a company that has figured everything out. You are building the category from the ground up — with real operators, real problems, and real consequences.
Hard systems problems with operational consequences
The problems here — signal pipelines, demand orchestration, AI-era content infrastructure — are technically hard and operationally meaningful.
AI-native product development, not superficial AI wrappers
We are not adding AI to an existing workflow tool. We are building infrastructure where AI is the core operating layer.
High ownership in an early team
Early team members own significant parts of the system. There is no large engineering org to disappear into.
Real operator workflows, not demo-only product surfaces
Everything we build is designed to work in production environments with real leasing teams — not to look good in a demo.