About NevaNeva Technologies builds AI tools and engineering support for the US construction industry. Our first customer is a US general contractor doing ~$250M a year in work. We are early stage, and building the first versions of our products directly against their real data: submittals, drawings, estimates, pay applications, and project files.

What you'll doYou will be an early engineer. You take a workflow from a project manager or estimator, rapidly build a custom AI tool from it, ship it, and keep it running. Each build has a user, a dataset, and a number that tells us whether it worked. The first few on the list:

- A cost-estimating tool that reads historical project data and a new set of plans and produces a rough budget with its assumptions stated.
- A pay-application checker that compares a subcontractor's monthly invoice against the schedule of values, prior invoices, lien waivers, and stored-material backup, and returns a discrepancy list.
- Be comfortable working with parallel teams building physical AI applications (~12 months after start).
- Claude skills and plugins that package our customer's working rules so every team gets the same result.
- Integrations with the tools contractors already use: Procore, Bluebeam, Microsoft 365.

You will talk directly to the people who use what you build, including the customer's executives,



and you will be expected to tell them what the tool gets wrong. Must have
- 3+ years building software, with Python as your main language and enough full-stack experience to ship a small web app on your own.
- Have shipped an LLM-backed tool to real users, including tool use or agents, and kept it running afterward.
- Can extract and structure data from PDFs, drawings, and messy spreadsheets.
- Build evaluations as a habit: a test set and a number, not "it looked right."
- Can deploy and operate what you build: cloud hosting, scheduled jobs, logging, and fixing it when it breaks.
- Have worked against third-party APIs and can judge whether an outside AI tool is positive enough to use.
- Can explain what you built, and its limits, to a non-technical person.

Nice to have
- Construction, civil engineering, or AEC software exposure (Procore, Bluebeam, Revit).
- Experience as a forward-deployed or customer-facing engineer.
- Team management and hiring experience
- Comfortable living in Excel.

What you get
- You set the engineering bar for a team that will grow around you in 2027.
- Your work goes into production for a real US contractor within weeks, not quarters.
- Direct exposure to how an AI consulting and product business gets built from the first customer onward.

📌 Applied AI Engineer (Bengaluru)
🏢 Neva Technologies
📍 Bengaluru