Software Engineer - Machine Learning/Agents and LLMs

About Us

Letterdrop is a YC-backed seed stage startup. We help sales-led companies uncover net new pipeline by automating social selling and giving them unique intent data. We effectively sell pipeline and we’re pretty darn good at it for the right fit companies (sample 1, sample 2, sample 3, and many more success stories). Everyone is using AI to create more spam. We’re zigging when others zag to help businesses build trust with their customers and win real business. You’ll be working closely with me, the CEO to design and build features.

About You

• You have strong CS fundamentals and background in ML and applying LLMs. You understand basics of memory, parallelization, runtime from a practical standpoint. ie. you don’t do things like create arrays with 60,000 elements and running out of memory. You are also very familiar with using LLMs in production.
• You are excited about being the founding ML engineer at a small startup. We have extensive ML capabilities in our product. We use technologies like Langgraph, OpenAI, Haystack, BigQuery, Cloud functions, Flask. You will be taking over as the first ML engineer.
• You are independent and learn quickly and know how to find out answers. Most of what you need to know is online in public documentation, GitHub Issues, Open-source code, forums, and our codebase. You can read, understand, and navigate through all of this yourself without handholding.
• You take pride in writing clean and maintainable code. You work on a team. You read others code. They read yours. Make both of your lives easy by making code simple and easy to understand. Do you hate yourself? No? Good. Then write code that you would want to read.
• You move fast. We’re a startup. We ship features quickly. You’ll be working through multiple bugs, improvements, and features per week.
• You are interested in building great product experiences. Our product is used by marketing teams at fast growing companies. If they run into bugs, poor UX, latency, or things flat our crash, that reflects poorly on us. You want people to use our product and go “Wow!”
  
  

Who is this for?

• You want to cut your teeth as an early engineer on a small but hardworking team.
• You’re not one for defined roles. You’re excited to work in conjunction with our full stack engineers to make things work.
• You’re excited to experience a lot of growth. Our engineers say that they’ve shipped faster here than any other job before. They feel confident from all the programming they do across different surfaces. It’s a great career accelerator.
  
  

Who is this not for?

• You can’t work hard right now. If you have other priorities outside of work, being an early engineer at a startup is probably not the right move.
• You need a traditional environment and structure. I’m a prior Product Manager at Google and technical founder. I care deeply about product and code quality. But I can’t handhold you since I also have to do sales, marketing, customer support, ops, HR, and a whole bunch of things. I meet with the team daily to do code reviews and answer product questions. Beyond that, you can ask your peers, but there’s no manager who’s going to sit by you.
• You’re not excited about what we do. We are building B2B software to help companies automate sales and marketing. Lots of working with customer data, LLMs, integrations, and some distributed systems.
  
  

What will you do?

• Architect and implement LLM/ML based features for our platform.
• Create infrastructure and tooling for handling high volumes of data in. Performance matters.
• Work with latest in AI and LLMs in an applied setting.
• Collaborate with team on code reviews and best practices to grow a scaleable codebase.
• Work directly with CEO to roadmap new features.
• Our stack is Flask, React, Node JS, Next JS, MongoDB, BigQuery, Langgraph, and others.
  
  

Requirements

• 1+ years of ML engineering experience.
• Comfortable with LLMs in a production setting
• Experience with production machine learning. You are comfortable with Python and agentic tooling like Langgraph, etc. It’s changing fast, so more important that you can learn quickly on your own.
• You have publicly viewable side projects on GitHub and can pick-up new technologies and frameworks quickly. Please share your GitHub.
• You are comfortable working on backend in addition to ML. We’re small so engineers have to wear multiple hats.
• Ability to act independently, work well with a team, and ship features quickly. A positive can-do attitude and confidence that you can solve any problem is needed at a startup where we're figuring things out as we go.
• Good English communication skills and ability to read documentation online to research how to solve problems yourself.
• Loose familiarity with Node, JavaScript, React, and MongoDB. It's ok if you're not, but as long as you're a quick learner and think you can pick up our stack within a month or two, you'll be fine.
  
  

Why you should apply?

• Lots of ownership on a small team.
• Directly interact with CEO to develop new features.
• You are looking to learn rapidly. This isn't a big company job tweaking small pieces of a giant project without impact. You will be shipping new features every week.
• We’re grounded: Too many startups raise lots of money and are building fluff. We’re pragmatic and seek to create value for our customers that they're willing to pay for. We’re break even and optimistic that we're building a platform that will be the de facto tool for thousands of B2B teams in 2-3 years.
  
  

Interview Process

• 15 min initial phone interview for introductions
• 60 min live coding interview on Zoom
• 2 hr take home project
• Reference checks from 1-2 previous mangers
  
  

We try to keep it as short as possible to be respectful of your time.