About Uplevyl

Uplevyl builds AI-powered knowledge and community infrastructure for organizations serving women.

Our products include UpGenie (our domain-specific AI assistant), WeHub (our community platform), and UpSocial (a social platform for women).

We hold a straightforward conviction: in high-stakes domains like rights, law, and financial security, a generic AI is not enough. The answers people stake their lives and livelihoods on need a purpose-built system with verified, native data. That is what we build, and it is why the work is urgent.

We have also made a deliberate choice about how we build: a small team of exceptional people, paid well above market, each doing work that would normally take several. We would rather be ten people who move the world than thirty who move paper.

About the role

Uplevyl is a data and technology company. What we sell is a knowledge layer: every verified US law on a subject, across every jurisdiction, held as one structured record per law that carries the jurisdiction, the right, the eligibility rules, the citation and the dates.

The first knowledge layer, workplace rights, is built and running. Survivor rights, the hardest subset of it, was built first and is live: four areas of law, sourced and verified across federal, state and local law. The rest of workplace rights is being populated on the same pipeline, wealth life events has a live prototype, and reproductive health is in R&D.;

The platform runs as five layers. The source layer finds, fetches, reads and checks the law. The data layer holds the versioned, jurisdiction-resolved corpus. The AI layer answers questions from that corpus through a seven-step agentic workflow. The enterprise layer gives each client its own instance, so the core never knows the tenant. The analytics layer reads all four and writes to none.

Here is the honest version. That diagram is true in production for survivor rights, and it is held together at the seams by spreadsheets and people. Source versions and metadata still move between the legal team and engineering as workbooks. The categories the assistant sorts a question into are fixed in a prompt rather than read from the data, so when the legal team adds one, the system cannot follow. Each layer has an owner. The seams between them do not.

You will own the seams, and the data layer they all run through. Your job is to make the diagram true in the system: one record schema for every law in every subject we build, versioned so the corpus can answer what applied on a given date, resolved to every one of the 91,000+ US jurisdictions, and served to the AI layer with the metadata it needs to be right.

You will do this hands on. This is a small team, and the person who draws the schema also writes the migration. You will work daily with our AI engineer, our source extraction engineer, the product and engineering pair building the source management portal, and the legal team that verifies every law we hold. You report to the India Team Lead and you will be in the room, on a whiteboard, with our founder when architecture decisions are made.

What you'll own

- The record schema. One schema for every law in every subject we build:



jurisdiction and level of government, category and right, citation and verbatim text, eligibility, enacted, effective and recorded dates, preemption status, provenance and version. You decide what is universal and what belongs to a subject, and you make the legal team's categories, down to the sub-category, live in the data rather than in a prompt.
- The versioned corpus. Raw, cleaned and current, every version kept. Changed laws are re-fetched on a schedule, diffed line by line, and pass through a review queue and a merge gate before they touch the live corpus. A struck-down ordinance stays, marked, and is never served as live law.
- The jurisdiction map. Every US jurisdiction, federal, state, county and city, resolved to the law that applies there or to an explicit record that says none does. The list of 91,000+ jurisdictions exists; resolving each one is the work.
- The contracts between layers. The source layer releases verified records; you define the shape they arrive in and the quality gate they pass. The AI layer retrieves from the current corpus; you define the index, the filters and the metadata it retrieves on, with our AI engineer. Analytics reads the corpus and the platform's anonymous logs; you define what it reads and make sure it never writes back and never sees a client's records.
- Tenant and subject inheritance. One shared record schema across clients, a separate instance and data store per client. A new client is a configuration record and a new subject is a new corpus; you keep both true as they multiply.
- The end of the spreadsheet as a system of record. The source management portal exists to check sources in and out, hold immutable versions, and connect straight to the pipeline.

Its data model is yours: the schema, the migrations, and the day the last metadata workbook is retired.

Who we're looking for We hire across levels, but this is not an entry point into data work. Systems you have designed and run in production matter more than years, and we will ask you to design in our process.

- You have designed a data model that other teams built on. A relational schema (Postgres or equivalent) that survived new domains and new clients, and that you migrated without losing history.
- You have built versioned, auditable data. Effective-dated or bitemporal records, append-only history, line-level diffs, review-and-merge gates. You know why "what is current" and "what applied on that date" are different questions, and you have made both cheap to answer.
- You have owned pipelines end to end. Ingestion with quality gates that catch a false success upstream, transformations that are reproducible, and monitoring that tells you a count is wrong before a user does. Python and SQL are daily tools, not occasional ones.
- You know what a retrieval system needs from data.



You have worked alongside an AI or search team and understand that chunking, metadata filters and re-ranking are only as good as the records underneath.
- You are precise about multi-tenancy. Tenant isolation, shared schemas with separate stores, and what must never cross a boundary.
- You use AI to build, and you check its work. Claude Code, Cursor, whichever: you use these tools daily, you review what they produce line by line.
- You write decisions down, plainly. Schema choices, contracts between layers, what changed and why.
- High agency. Nobody has held this role here before. You will find spreadsheets where a system should be, and you will decide the order in which to replace them.

Even better if

- You have modeled legal, regulatory or statutory data, or any corpus where jurisdiction and dates decide the answer.
- You have built or run a source-of-truth system with human review in the loop.
- You have worked with graph databases, or with graph-shaped relationships in a relational model.
- You have shipped on Google Cloud or Vertex AI, or a comparable managed cloud AI stack.
- You have worked in mission-driven, social-impact, women-focused or nonprofit ecosystems.
- You have founder or startup operator experience.

Before you apply You will inherit a live product with real users and real partners, a corpus that a legal team and engineers spent months verifying, and a set of spreadsheets doing jobs that a system should do. The first months are as much archaeology as architecture: finding where the diagram and the database disagree, and choosing what to fix first without breaking what works. If you want a greenfield build with no history, this is not the right fit, and that is fine.

It is a hands-on role of one, on a small senior team, in our Noida office. The work is demanding, the expectations are real, and the rewards, in compensation, in ownership, and in what you will learn, match them.

If you would read about bitemporal data models on a weekend anyway, you will be at home here. If that sounds like a chore, we would rather you knew now.

If you want to be the reason a system that answers questions about people's rights can prove where every answer came from, we want to hear from you.

Our interview process is thorough, and we will walk you through every step of it in our first conversation.

How we work at Uplevyl

- We finish what we start, and we do it well. We follow through on what we say we'll do, and we care about the difference our work makes.
- We listen before we decide. Before we act, we ask who it actually affects: a customer, a partner, a teammate, or the communities we serve.
- We move with ownership and urgency. We don't wait for perfect information or for someone else to raise their hand. We make the call, we move fast, and we hold a high bar for quality.
- We are better together than alone. We work across teams, say what we actually think, and go out of our way to help each other succeed.
- We stay curious and keep learning, including how to use AI well. We hold ourselves to the bar we're building toward: questioning our own assumptions and using AI to think better and move faster.

📌 Data Architect (Noida)
🏢 Uplevyl
📍 Noida