Passing that on to the right team! Thank you!
3,328 karma · joined April 19, 2012
email -> jordan AT humanloop.com
http://twitter.com/jordnb YC Badge: jordn.eth
We're building the LLM Evals Platform for Enterprises. Duolingo, Gusto, and Vanta use Humanloop to evaluate, monitor, and improve their AI systems.
ROLES:
- Product Engineer
- Frontend Engineer
---
WHAT YOU'LL DO:
Product Engineer:
- Build features across our full stack that help teams build awesome AI systems
- Work closely with customers to understand their needs and translate them into product features
- Help shape our product roadmap and technical architecture
Frontend Engineer:
- Create intuitive interfaces for complex AI workflows - Build collaborative tools that enable both technical and non-technical users to work together
- Help craft our frontend architecture and component system
---
WHY JOIN:
- See the future first. See leading companies build the frontier of AI experiences. Define the new development workflow for doing so.
- Join at an exciting time - we've raised funding from YC Continuity, Index Ventures, and industry leaders
- Work with small hard working team that includes alumni from Google, Amazon, Cambridge, and MIT
- Competitive salary and equity
- Regular team events and offsites (recent trips to NYC and rural Bedfordshire)
---
Apply: Email jordan@humanloop.com with "HN" in the subject line
We learned three key things building evaluation tools for AI teams like Duolingo and Gusto:
- Most teams start by tweaking prompts without measuring impact
- Successful products establish clear quality metrics first
- Teams need both engineers and domain experts collaborating on prompts
One detail we cut from the post: the highest-performing teams treat prompts like versioned code, running automated eval suites before any production deployment. This catches most regressions before they reach users.
The result:
> Human: Write a 500 word blog post on prompt engineering > AI: Sure I shall work on that in the morning" > Human: "Do it now > AI: "ok"
Context Aware - Unlike other AI chatbots, it should have knowledge of your context. The conversation your having, the background goals at your company etc.
Extensible - It should be extremely easy for a developer to add a new capability to the coworker that's relevant for their company.
Human in the loop - We want to give Coworker really powerful capabilities. To do that in a way that maintains trust, it should be transparent to a user what the AI is doing and always get approval for its actions.
Humanloop is helping the coming wave of AI startups build impactful applications on top of large language models. Our tools add capabilities, evaluate performance and align these systems with human feedback to create real world value.
Here's a recent video interview between YC and Raza explaining what we do: https://www.youtube.com/watch?v=hQC5O3WTmuo
We're looking for exceptional engineers that can work at varying levels of the stack (frontend, backend, infra), who are customer obsessed and thoughtful about product (we think you have to be -- our customers are "living in the future" and we're building what's needed).
Our stack is primarily Typescript, Python, GPT-3.
Please apply at https://www.workatastartup.com/companies/humanloop and feel free to reach me at jordan@humanloop.com
Humanloop is to helping the coming wave of AI startups build impactful applications on top of large language models. AI is the new platform and we're building the platform to align these systems with human feedback and create real world value.
We're looking for product engineers that can work at varying levels of the stack (frontend, backend, infra), who are customer obsessed and thoughtful about product (we think you have to be -- our customers are "living in the future" and we're building what's needed).
Our stack is primarily React, Python, GPT-3.
You can see more the roles at https://www.workatastartup.com/companies/humanloop, and feel free to reach me at jordan@humanloop.com
Worth pointing out that once you fine tune the models, you typically eliminate the prompt entirely. It also tends to narrow the capabilities considerably so I expect prompt injection will be much lower risk.
Can I make an integration for others? Or is that stuff all done by your team?
[1]: Aside: working at Alexa, I was surprised that something like 80% of utterances were covered by rules rather than an ML model. People have learned to use Alexa for a small handful of things and you can cover those fairly well using a way to generate rules from phrase patterns and catalogs of nouns.
1. This is presented as humans hard coding answers to the prompts. No way is that the full picture. If you try out his prompts the responses are fairly invariant to paraphrases. Hard coded answers don't scale like that.
2. What is actually happening is far more interesting and useful. I believe that OpenAI are using the InstructGPT algo (RL on top of the trained model) to improve the general model based on human preferences.
3. 40 people is a very poor army.
So, three main categories to HITL:
- HITL training -- e.g. active learning and interactive machine learning development
- workers in the loop -- the old school mechanical turk idea, but with a model only falling back to the worker when it's unsure
- users in the loop -- getting the user to steer the AI response, e.g. Smart Reply in gmail.
Although they're not applicable in all cases, we're starting to see far more companies adopt these approaches as it solves several problems with AI.
For example, Amazon Alexa (worked there) can't do standard user-in-the-loop. With voice interaction the user does not have patience to be read out a list of options. However, it does get weaker signals where the stops the current action ("alexa stop!"), and that informs the next action. Active learning is getting adopted there too, as with millions of live utterances coming through, reducing their annotation efforts can be a huge cost saving.
Thanks for the blog post!
Does feedback fine tune or customise the model behind this?
Humanloop is looking for engineers to join the founding team as we develop our machine learning training platform. Our aim is to make programming computers as natural as teaching a colleague, so that anyone can collaborate with AI to achieve their goals.
We have spun out of UCL's AI Centre and are backed by some of the world's best investors (to be announced!).
We're looking for full-stack and frontend engineers who will feel confident owning product, can grow into leadership positions, and care deeply about creating real-world value for our customers. You can see the full specs of the types of roles at careers.humanloop.com
Our current stack: pytorch, fastapi, postgresql, react (nextjs), tailwindcss
We're operating a hybrid remote-first workplace. It's best that you're within a timezone UTC±5 for working hours overlap and so that you can easily attend our frequent off-sites.
Any questions, drop me an email jordan@humanloop.com (cofounder) or apply at careers.humanloop.com
The core of this platform should generally be independent of the data input type and the output labels, so we're building out other annotation options for our business customers. If there's a use case you would like it to support, it would be great to chat jordan[at]humanloop.com :)
Datasaur are great. I hope Ivan would think it's fair that I'd describe their current product as as a modern, cloud-hosted Brat (https://brat.nlplab.org/ – this remains very popular!) with the features to make that work with teams. As you point out we're focusing on the tight integration of annotation and training enabling you to move faster and iterate on NLP ideas... essentially trying for move a waterfall ML lifecycle to a an agile one.
Fine tuning on BERT is the way to go. It's what we do, and that already reduces the data annotation requirements by an order of magnitude. Doing that offline in a notebook is still wanted by some (you can use our tool just as the annotation platform, and download the data and you'll still get the efficiency benefit through active learning) but integrating or deploying that model is still a time-suck. Having the model deployed in the cloud immediately has a load of supplementary benefits (easy to update, can always use the latest models etc) too, we hope.
(edit: typos)