Principal Software Engineer

Hey! Shun here, I’m the CEO and co-founder of Anima. We’re building an active learning OS for all of healthcare and life sciences towards maximising human wellbeing globally.

My entire life, I’ve been pulling on a thread that’s affected all of us in some way. Millions die every year because their medical problems aren’t treated quickly enough. Hundreds of millions suffer pain, worry and discomfort needlessly because of long waiting times. I trained as a doctor and quit out of frustration at seeing countless cases of people dying because they got misdiagnosed or didn’t get the right care plan. In one devastating incident, the healthcare system had simply forgotten to follow up about abnormal test results.

Pulling the thread led me to the health spending team at HM Treasury, responsible for deploying 100s billions into the NHS - there, I worked on workforce and system reform policy with No.10 and NHS England (most notably on forming statutory ‘integrated care systems’, which was signed into law as part of the Health and Social Care Act). Seeing the problem space at all abstraction levels convinced me that the only way to fix healthcare was to build a ‘Care Enablement’ platform that can automate and abstract away work at the clinic, and 10x doctors. Doing so would also be the path to a superhuman personalised medicine agent that could go well beyond humans, crunching tens of thousands of low level features at genome and transcriptome level.

At Anima, you’ll help us extend the 3 existing product lines we have, that millions of patients use, and build out new ones at the very cutting edge of healthcare reinforcement learning and agentic AI. Your work will save countless lives.

Top 1% growth - we grew 450% in 2024, are cash flow positive and at Series B/C metrics with a team of ~20

I won’t write too much here, because I think the metrics speak for themselves and if we get further along the process, we can always discuss it more. Users love our product. The biggest testament to this is that most of our growth comes from virality and referrals.

As one of the standout companies in YC’s W21 batch, we raised $2.5M before Demo Day in just 4 days, led by Hummingbird. In early 2024, we closed an oversubscribed Series A round of $12M, led by Molten Ventures. We’ve been turning down pre-emptive offers of more capital since 2024.

Anima clinics look after 3 million lives, and every day around 20,000 people get care through Anima, growing fast.

Why you might love working here

Since the beginning, we’ve cultivated a unique hacker culture at Anima with a flat hierarchy. If you look at some of our LinkedIn profiles, you’ll note that the title is often ‘Product at Anima’, which is a more accurate description of what your role will be - it’s much broader than the typical SWE role. We predict that small, killer T-shaped teams will win out over large mediocre ones in the era of agentic AI. We've been purposefully building a 'knights of the round table' where each person has an incredible amount of ownership, autonomy and human impact on saving lives - zero line managers, PMs or EMs, but an intensive focus on coaching and teaching. [In case you’ve seen the show, the best analogy is Firefly and the crew from it. Greatest show ever imo.]

It started with me. I self taught and wrote a lot of the Anima 1.0 code, and Anima’s active learning patent. I run most of the hiring tech chats to this day. I first and foremost see myself as an IC and builder, and still build stuff in my spare time. More than 80% of our product team are VPs, ex-technical founders and former CTOs (we do not have titles internally). We also have domain experts, including full-stack clinical engineers: ex-MDs who left established medical careers to join Anima because they recognised the much higher leverage they’d have here.

Almost everyone in Anima is at least somewhat technical and very UX focused. This makes for a truly flat hierarchy. Internally, people have described it as a hive mind - there is little semantic error because we all deeply understand the problem, engineering, grading for potential features, and trade-offs.

It's reflected in how we operate:

1. With the outlier product talent density at Anima, there's a heavy focus on coaching and teaching, with no need for line management.

   We've been able to make huge strides in the NHS because we all deeply understand the problem, grading tasks based on our engineering expertise, and real life clinical trade-offs. This deep domain knowledge is ultimately how we integrated with the major electronic health records (EMIS and SystmOne) in just 4 months, and grew from $150k to $7m in 20 months and 3 product lines.

2. We're relentlessly resourceful. In 14 months, we built an extremely loved enterprise app that users have said 'seems too good to be true'. Anima can take an information complete medical history as good or better than a typical human doctor, and increasingly automates front and back office in a way that would’ve seemed like science fiction a few years ago - for example, Annie (our AI) now deals with a large percentage of documents completely autonomously without a human in the loop.

3. Anima is a safe haven for free thinkers and we've been careful to build a culture where everyone feels comfortable being their complete unfiltered self, sharing their honest thoughts, feelings without ever needing to self-censor. Joining Anima means joining a crew with exceptional talent density, and your crewmates will never let you down.

4. We hire talented people who think from first principles and have high growth potential - great decision makers who deserve to have complete autonomy and are forces of nature when empowered with it. Decisions are never made in isolation by the founders. All information is transparently available to the whole team on our Notion - every meeting, decision, success, failure. We have a live Retool dashboard which shows live MRR, linked to Stripe invoices.

5. We are deeply collaborative, and through mutual and self-challenge, we converge towards the optimum, and decisively execute. We are united by child-like intellectual curiosity and experiment and wander freely when the right path isn’t clear. We believe all future managers should be formidable individual contributors & domain experts.

6. We hire candidates from all over the world. We offer flexibility over work schedule and location.

7. We have a good time! We get the whole team together at least twice a year for team retreats, with the last ones in Brittany (France), La Paz (Mexico), Tuscany (Italy), Dominican Republic!

Build With Us

Today, there’s a lot of hype around ‘verticalized AI’ but when our journey started in 2021, people outside of the ML community barely knew or cared about active reinforcement learning. ChatGPT didn’t exist. I remember getting frustrated and thinking… ‘don’t you get how big of a deal this is?!’

Since Day 1 (as written in our YC application), we’ve been building towards the holy grail of personalised medicine and deep phenotyping, powered by our proprietary active learning architecture. As I said earlier, I actually submitted a 2021 patent that was prescient for the sparse MoE and active learning loop that is commonplace today with LLMs. We’ve been building and building towards this secret plan this whole time, and now have one of the biggest, highest quality labelled datasets in the world.

When everyone is technical and make great decisions, it’s much easier to stay on the same page & execute rapidly. This means we have a super short latency from ideation to real usage.

Here are 2 concrete examples with specs and timelines (we’d be delighted to demo any of these to you):

• Alex, a clinical engineer, built ‘Slack for medical teams’, supporting 1000s of channels per organisation and a real time single source of truth for patient data, in 4 weeks, picking up a large set of new technologies on the way (ground up built from low-level services like Appsync, GraphQL, not Twilio).

• Dennis built a lightning fast cloud document library for clinics in a few days, with <100ms traversals through preloading. Recently, when faced with an ancient legacy API that took 30s to return an array of hits, he hacked together an async indexing service with caching that reduced latency to <2s without harmful race conditions. He built and deployed to prod this in under 24 hours.

If high growth delta and joining an elite scrappy crew is your priority, you’re going to love it here.

Does this sound like you?

• Hungry and wants their shot to change the world - a force of nature when empowered with the tools, resources and development to do it. Sees joining Anima as potentially their shot to do this, and takes duty to crew and mission extremely seriously.

• Obsessively concerned with UX, and optimises for this when building features rather than arbitrary technical goals.

• Bored and frustrated at big companies; feel held back by red tape, bureaucracy and poor decisions.

• Keen to understand the big picture & entire context of the company and vertical; impatient for growth towards a senior executive role.

• Expert competency in TS, plus being reasonably tech + language agnostic. Comfortable with key frameworks/libraries like Angular, Node and React. Able to work full stack in JS/TS. Values pragmatism and open discussion from first principles rather than dogma.

• Seeks to maximise not only self productivity, but combined team productivity, communicating the right things at the right time through the right channels (verbal/Slack/Notion).

• Disciplined towards best practice version control, CI/CD and code extensibility. Values ‘interface safety’ through dimensionality reduction at interfaces.

• Exceptional at ‘breadth-first search’ through Googling when tackling new challenges, and consistently mindful of local maxima.

• Intellectually curious with a growth mindset - able to tackle entirely novel challenges that lack prior precedent through applying strong CS fundamentals and first principles thinking, creatively using the right data structures & algorithms to solve problems 90/10.

• Familiarity with AWS (e.g. APIG, SQS, DynamoDB, Lambda, Cognito, Amplify, CloudFormation) and/or hungry to learn.

We don’t enforce any particular experience level, but you’ll need to demonstrate most of the above through past projects and/or our assessment process.

Our current stack & what to expect from the role

We are tech agonistic, and collectively choose the best tools for the job. We’re constantly looking to maximise our productivity and minimise what we call “discounted dev time cost” for shipping features.

We have 2 separate fully functional web apps in prod: one for clinical users and one for patients. Our stack is currently entirely in JS/TS: Angular + Capacitor + Electron, React (internal tools), Amplitude (analytics), a fully serverless backend in AWS (Cognito, Appsync GraphQL, Lambda, DynamoDB). We have good functional & unit test coverage and CI/CD.

Our stack is in a great place already: highly scalable, cost effective, good test coverage, easily maintained, secure and performant with minimal to zero ops. The product and codebase are stable and loved by our users. We write, test, deploy & ship new features rapidly.

We’re looking to add talented engineers who are hungry and understand the urgency and importance of what we’re doing for society.

First month - some examples of what to expect:

• Help add further key third party API integrations, including with legacy EMR systems and national APIs like e.g. the electronic prescribing service, allowing Anima to directly issue prescriptions.

• Iterate on a proprietary graph traversal algorithm to improve patient care and clinical value, and increasingly move away from explicit curation to implicit curation by NNs.

• Ship important features that will directly increase delta lives saved in your first 2 weeks.

• Join customer calls to develop a deep understanding of their fundamental motivations and needs/pain points.

Next 6 months - some examples of what to expect:

• Help architect and deploy a scalable & cost effective ETL data pipeline with version control, outputting clean data ready for tokenisation.

• Help deploy our active deep learning training & validation architecture to prod, so that we can correctly eat up our ‘1.0’ systems at the right time

• Build cutting edge products like global context aware chat with semantic search, care orchestration and LLM-enabled cloud telephony e2e

• Hire/scale the team, while implementing the right processes at the right times to maximise discounted team productivity and minimise discounted dev time cost for shipping.

6+ months - some examples of what to expect:

• Potential to transition to a more managerial/executive role. Lead an autonomous lance of elite engineers to fix healthcare and save lives.

• Work with the ML/data team to creatively ideate and ship features to improve ETL pipeline throughput and quality through a data-driven approach powered by analytics.