Vectorview
Building custom evaluation tasks for AI
Software Engineer
About the role
As part of our founding team, you will have a high agency role in building the Vectorview framework for AI evaluations: shipping new features, talking to users and shaping company culture for other team members to come.
You will work directly with the founding team to primarily iterate based on customer feedback, help with infrastructure scaling (cloud and on-premise) and improve our proprietary evaluation framework.
Who’s a good fit?
We think you should apply if the following sounds like you:
- Proactive — You find problems and independently initiate a solution
- Productive — You get a lot done quickly while having a high bar for quality and craftsmanship
- Hard-working - Strong work ethic and willing to go the extra mile
- Fun – You’re fun to work with and can lift up others even when things are not going well
What we’re looking for?
In addition to the personal qualities mentioned above, this role requires the following:
Must haves
- 2+ years of software development experience (side projects and internships count)
- Excellence in Python
- Experience with LLMs (OpenAI API, Langchain, HuggingFace, etc.)
- Demonstration of autonomy and entrepreneurial spirit (side projects, previous companies)
- Machine learning knowledge
Nice to haves
- Proficiency in Linux, cloud deployment, CI/CD
- You’ve led a technical team
- Previous startup experience
- Full stack experience (infrastructure, backend, frontend, testing)
- Published an ML paper
About Vectorview
Vectorview works with the big AI labs to evaluate the capabilities of their foundation models and LLM agents. Our evaluation framework is used to easily create, run, and score custom evaluation tasks. This is the key in creating safe, robust and reliable models. For example, we enable foundation model companies to audit dangerous capabilities in the next generation of models.
We’re creating the new standard for how to evaluate the agentic capabilities of LLMs. Foundation models will become as ubiquitous as javascript and our evaluations will be seen every time a company publishes a new model.
Our passion for AI safety is based on our belief that AI will transform our world for the better—but this won’t necessarily happen by default.
We’re at an early stage in our journey and becoming a part of the team now is a once-in-a-lifetime opportunity to create something new that will have a massive impact on the future of AI.
