AI/ML Engineer ( leadership role)
We're Continuum Tracker, a European AI tech startup focused on creating the first AI aimed at recommending product features worth building and ultimately helping more products succeed (more on our tech stack below).
How our platform works
We predict (= recommend) product features for any product based on its business goals. Users can choose which of our recommendations to build.
After a user selects a recommendation, our AI creates a synthetic understanding of the user problems that drive the feature’s desirability, along with potential solutions.
Lastly, this intelligence is passed on to an AI-assisted coding agent that can ingest the data and build the solutions. More on that here.
Ultimately, our users build more desirable features that solve more valuable pain points-directly connected to their business goals. Or use our recommendations to automate prioritization and speed up the building process, saving money.
Your role as leader
Making these recommendations is a tough task for AI. It requires AI to understand both the market the product operates in and the product itself.
It’s a hard problem. That’s exactly why we’re building it. You’ll own the engine-and what comes out of it . If you have doubts about why this might work, read our article on the product development cycle automation.
Little bit about the market
There are 8 million product builders worldwide without product-led growth experience on Lovable, and 1.6 million product managers across companies worldwide. We help builders succeed and PMs move faster.
What will you do:
Own the engine’s outputs -shaping the technologies, methodologies, and direction of the AI pipeline (LLMs, advance statistics, machine learning) alongside another AI engineer and the Founder with help of data labelers.
Build the methodology and technology for predicting the likelihood that a product feature will be built.
Design and scale AI evaluations to measure the quality of our product recommendations.
Drive AI cost efficiency while maintaining output quality.
Solve the scaling challenges required to process 1M+ products for each new recommendation*.
Optimize search and retrieval to identify similar products across multiple criteria.
Run deep-dive analyses to uncover new approaches for improving the AI engine’s accuracy, scalability, and performance.
Build a data foundation that enables the platform to scale seamlessly across Europe.
*There is 1 billion active products in 2026 to give you perspective.
What you might find during your work
Some parts of the AI engine may be a great fit for fine-tuning or developing smaller LLMs for targeted tasks such as data extraction. Depending on your expertise, you’ll help us explore these directions where they make sense. Important consideration, swe have limited compute resources, so we need to be thoughtful and strategic about how we build.
Our tech stack:
Python · FastAPI framework · Google Infrastructure · Multiple cloud LLMs · A few models running on server in cloud · Qdrant as vector database · LLMs as judge methods · Multiple prompts to optimize · Claude as main DEV tool with our AI workflow · Label studio for data labeling · GitHub · Digital Ocean · ReRanker · OpenAI whisper · MCPs · LangSmith · Langchain · Sentry · PostgreSQL · Our own context cost management tool · Agentic architecture in general.
You're a fit if you:
Have deep AI expertise and understand which AI techniques work best for different tasks.
Have strong hands-on experience building AI systems
Understand architectural concepts such as reranking, vector databases, neural networks, fine-tuning, model evaluation, advanced statistics, and agentic LLM system design.
Have the drive to build toward our vision of helping more products succeed by building the largest database of products and automating their development.
Have a growth mindset and is motivated by ambitious goals.
Are comfortable working asynchronously and clearly communicating and presenting your work.
Use Claude or similar AI coding tools daily.
Have very good English in speaking, reading, and writing.
Understand that this is not a highly paid opportunity.
Nice-to-haves
US start-up experience
Loves to celebrate small victories
Ideally have built and scaled similar recommendations system
Be able to pitch on stage to non-technical people
Experience with data labeling and annotations
You did not find yourself in those criteria? Let's meet anyway.
Why join
We are building something unique, be part of the story.
We are privacy-focused. If you wish not to be named on the internet due to privacy concerns, we respect that.
We do not need a full-time commitment. If you have spare hours in your day-to-day work, that is perfect for us.
Except compute, there is almost no limit on the technology we allow.
Be part of a small team who is preparing to scale to Europe.
Work anywhere from the world (EU preferred).
If you take on this challenge, you might go with us to start-up events as speaker or attendee. This depends on the engine's performance to be honest
This entire text was written by a human , as a small gesture of appreciation for the hard work and dedication of everyone who joins us.
Image shows how we were competing in Warsaw on Start-up competition show
Let's meet to see potential cooperation
Meet Tomáš Veselý at some lovely cafe in Prague or over the Google Meet. Let's talk about our product. Some question we are really interested:
How would you approach our problem if you were building the system from scratch?
Imagine a hypothetical scenario: your company is mostly engineers, with limited product and business expertise. How would you decide what to build next without years of product or growth experience?
What does your ideal way of working with a remote team look like?
What would motivate you to take on the challenge of building an EU-first LLM for product recommendations? Do you have relevant experience, or would you be interested in developing the expertise needed?