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 work with mission-driven partners to turn complex, high-stakes information into clear, trustworthy guidance, and to build systems that create lasting impact.
We hold a simple 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. AI is reshaping how the world learns, works and earns, and the women we serve cannot afford to be left behind.
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. That choice sets the bar for every hire, including this one.
About the role
The AI Engineer, Retrieval and Search Systems, owns the retrieval layer underneath our AI products: how our documents are chunked, embedded, indexed, searched and ranked, and the open models that run on top of it. You will build this system on our own infrastructure, improve it toward a small language model (SLM) trained on our own data, and maintain it as our content, users and models grow.
This is hands-on implementation work from day one. Chunking and embeddings need continuous care: as new documents arrive, formats change and models are upgraded, they have to be re-tuned, re-indexed and re-evaluated. You will own that work on an ongoing basis.
You will ship to production from your first weeks, with real users and real partners depending on what you build.
What you'll own
- Chunking. Chunking strategies for long, mixed and structured documents, and re-chunking as our content and sources change.
- Embeddings. Selecting, running and upgrading open embedding models, and managing re-embedding whenever models or content change.
- Vector indexing and storage. Building, tuning and scaling the index, and keeping it fresh as documents are added, updated and removed.
- Search, retrieval and re-ranking. Hybrid lexical and semantic search, re-ranking, filtering and relevance tuning across our document collections.
- Model serving. Running open models (Llama or similar) on our infrastructure, with consistent tokenization and context handling across models, at a latency and cost we can sustain.
- The path to an SLM. The data pipelines, fine-tuning setup and evaluation needed to train and run a small language model on our own data.
- Evaluation and monitoring. Test sets, retrieval metrics, regression checks on every change, and production monitoring for drift and quality.
Completion of the project
The initial build is complete when:
- The full retrieval pipeline runs in production on our own infrastructure.
- It meets the quality, latency and cost targets we agree on our test set.
- An open model is served on our infrastructure on top of it.
- Evaluation, monitoring and documentation are in place.
After that, you own the system: maintaining and improving it, and leading the move toward our own SLM.
Only apply if
You have hands-on experience in all four of these areas, and you have worked on at least one previous project that involved them:
- Search and retrieval
- Embeddings
- Chunking
- Vector indexing and storage
If you are missing any of these, this role is not the right fit right now.
Who we're looking for
We hire across levels. Depth of understanding and results you can show matter more than titles or years of experience. We will ask you to build during our process, because we want to see how you work on a real system.
- You have built retrieval systems that run in production. You can walk us through a system you built end to end, the choices you made, what broke and how you fixed it.
- You ship working code. You are fluent in Python and build production services with Docker, Kubernetes and GPUs. Your code is tested, documented and easy for others to run.
- You measure before you claim. You use test sets and metrics such as recall@k and nDCG, and you check every change for regressions before it ships.
- Rigor. In our domain, a confident wrong answer can harm someone. You test before you claim, and you flag what you are unsure about.
- High agency. You unblock yourself. Given an ambiguous problem with no owner in sight, you scope it, decide and build.
- A clear communicator. You can explain a technical trade-off to a non-technical partner and defend a design choice to a senior engineer, in plain words either way.
Even better if
- You have served open models in production with tools such as vLLM or TGI.
- You have fine-tuned or trained small language models, or built the data pipelines for them.
- You have built retrieval over legal, regulatory or other high-stakes documents.
- You have worked with cloud AI platforms on GCP, AWS or Azure.
- You have public work we can look at: open source, technical writing or talks.
Before you apply
We want to be direct about what this is. This is a high-bar, full-time role on a small team, with real ownership and high expectations. You will build the system, improve it and keep it running, and much of the value comes from the ongoing work after the first release. If you want a pure research role, or a role where you advise and others build, this is not the right fit, and that is fine.
If you want to build and own the retrieval system behind products people depend on, we want to hear from you. Our interview process includes a hands-on build task, 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. Trust here is built the plain way, by consistently showing up for people.
- We move with ownership and urgency. We don't wait for perfect information or for someone else to raise their hand. If something looks like it could go wrong, we say so early. 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. Leadership here is measured by the impact you create.
- 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.

