01 · Long horizon
Work on a foundation that can compound.
The durable advantage is not a one-off prediction. It is infrastructure that makes biological information more useful as research advances.
Careers · Point Eight AI · Remote or Singapore
Build what the life sciences will depend on.
We are an early team working across the life sciences, machine learning, data, and product. The problem is hard enough to require deep specialists — and connected enough to reward people who cross boundaries.
§ 01 · Why this work
Ambition, with scientific restraint.
We want to build a consequential company without confusing a compelling demo for evidence.
02 · Product depth
Make complex research legible.
Product work here means shaping how scientists and models share state, compare evidence, resolve uncertainty, and decide what deserves the next expensive step.
03 · Technical honesty
Evidence before claims.
We distinguish implemented capability, planned architecture, computational hypothesis, and experimental proof. That precision is part of the product.
04 · Early team
Own the whole shape.
At this stage, research choices affect product choices, and product choices affect the company. People here have a real hand in all three.
§ 02 · Current conversations
Where we need leverage.
Titles are starting points. We care more about the problem you can own, the evidence you use, and the quality of judgment you bring.
The immediate priority is product strategy: turning a deep technical direction into a focused sequence of users, decisions, and proof. We are also building the research and model layers around it.
Product Strategy
Give the science a product shape without flattening it.
Own the research-to-product thesis, user discovery, programme sequencing, and the boundary between workflow, models, services, and partnerships. Strong writing and first-principles judgment matter more than a familiar playbook.
Immediate priority
Early team
Start a conversation →
Research Systems Engineer
Build infrastructure that can survive scientific scrutiny.
Work on typed research state, immutable lineage, model adapters, verification, artifact systems, and interfaces for human review. You should enjoy making complex workflows explicit and difficult to misuse.
Engineering
Python / systems
Start a conversation →
Biological ML Researcher / Engineer
Move from generic representations toward useful biological decisions.
Develop and evaluate biological models, design honest baselines, connect representations to measured outcomes, and make uncertainty visible. We value rigorous negative results and reproducibility as much as a new model idea.
Models
Evaluation
Start a conversation →
Founding Collaborator
For people who want to shape the research programme and the company together.
We are open to experienced scientific, product, and company-building partners whose work is unusually aligned with this direction. We prefer substantive mutual diligence and a small piece of real work before deciding what form the relationship should take.
Collaboration
Scope shaped together
Start a conversation →
Open Application
If the fit is obvious to you before it is obvious to us.
Computational biology, scientific operations, applied research, design, partnerships, safety, or another discipline we have not named — send the work that best shows how you think.
Open
Remote / Singapore
Start a conversation →
§ 03 · How we work
Make the question clearer. Then make the system earn the answer.
01 · Frame
Start with the decision.
Define what has to be learned, which evidence would change the path, and where authority sits.
02 · Build
Make state explicit.
Encode inputs, constraints, provenance, and handoffs so another person can inspect the work.
03 · Challenge
Look for the missing edge.
Test baselines, failure modes, leakage, hidden assumptions, and where a score overreaches.
04 · Learn
Carry evidence forward.
A result changes the next round without erasing the record of how the previous one was reached.
§ 04 · The conversation
Mutual diligence, through real work.
We keep the process compact and adapt it to the seniority and shape of the relationship.
Send context
A resume is useful; a piece of work, a sharp memo, or a short account of what draws you to the problem is better.
Talk through the thesis
We compare how we see the market, the science, the missing evidence, and the most important next decision.
Work a real question
We use an existing problem or prior work, not generic puzzles. The point is to see how judgment travels between us.
Shape the relationship
Role, scope, location, and pace follow from the work. We make the mutual expectations explicit before either side commits.
§ 05 · Start here
Bring the work that shows how you think.
Tell us what you would want to make true.
Send a concise introduction and the strongest example of your work. If you are exploring a partnership rather than a role, say so plainly — both conversations are welcome.