Founding Engineer Role Inquiry

preceeding: [Hiring] [Remote] MIT AI startup, Founding Engineer Want to build the future of better, more affordable healthcare? We're an a16z-backed, MIT-founded startup with engineers from Apple & Microsoft, using machine learning to transform how companies design and manage healthcare. Our customers include public companies & top 5 insurance brokerages, already seeing major gains in efficiency & cost savings. What we’re looking for: • Early-stage startup experience (pre-seed/seed) • Proficient in React, Node.js, Express, Firebase, GCP, REST APIs • Experience with AI agents (RAG, chaining, retrieval, etc.) • Familiar with secure infra & data best practices (encryption, access control, API/cloud hardening, etc.) If you're interested, DM me your LinkedIn/GitHub. ---- make this shorter in reply to the preceeding: I am interested in the role, however I do not know if I would be qualified. I have experience with react, node.js, next.js, minimal Express, no Firebase, little GCP, REST APIs I am mostly sure I know but who knows I may only have a basic understanding other than my experience building Django/Flask/FastAPI backends with Next.js frontends or sometimes just Vite, but almost always in TypeScript for frontend, unless I use Streamlit for a basic UI in python. Python is what I am most comfortable with. I really enjoy the application of higher math to solve problems and that is what I most enjoy about programming except I also understand and have build frontends that allow human in the loop and reinforcement through human feedback. Such as this current application I am building. Central to it is a PersonaAgent Class: class PersonaAgent(Agent): def __init__(self, name, role, traits, system_prompt, memory=None): super().__init__(name, role, system_prompt, memory) self.traits = traits # e.g., {"tone": "sarcastic", "vocabulary": "high", "curiosity": 8} def build_prompt(self, input_text):d persona_details = f"You are a {self.role} with the following traits: {self.traits}." return f"{persona_details}\n{self.system_prompt}\n\nUser: {input_text}\n\nAgent:" Except instead of simply using a dictionary of keys and qualitative values you instead use a number between 0 and 1. This allows you to easily manipulate the values and store them in a sqlite database or since they are just numbers you could use a vector database such as Chroma. Why would you do that? Because it allows you to have an object of permanence that is the "persona" that you can then use to shape the LLM call. Not just that but these values can be manipulated through whatever logic you can think of, such as using real time data from RSS feeds, fetches or any number of other tools you can define in the PersonaAgent class above. So I used all of this to create github.com/kliewerdaniel/news17.git which is my unfinished work in progress. I have built a basic RAG so I understand how that works, just not something good or at scale, yet since it was just to teach me the concept, you can see my blog post here: https://danielkliewer.com/blog/2024-12-01-Basic-RAG You might like this book I wrote about Agentic Knowledge Graphs: https://6340588028610.gumroad.com/l/ddsrtm As far as being familiar with secure infra & data best practices (encryption, access control, API/cloud hardening, etc.) I am a complete novice, I am sad to say, so I do not even know if I would be qualified. My github is github.com/kliewerdaniel/ My website is danielkliewer.com My linkedin is not really related to development but rather art and data annotation so I don't know if that would be helpful. I do have a decade of experience with data annotation if that matters so I know the pipeline pretty well. I took classes from edx.org and ocw.mit.edu for free which taught me the math and computer science I have learned about AI. So I may not be qualified, you would have to see my github to see, I am but a hobbyist and do not make things to ship so I have almost no experience with your last requirement about secure infra, as all I really know is how JWT works and use some basic attempts to stop XSS but honestly I don't really know what I am doing.

Here’s a shorter, cleaner version of your message that still conveys your experience and enthusiasm, while being honest about your current gaps: --- Hi, I'm interested in the Founding Engineer role—though I’m not sure I meet all your criteria. I’m strongest in Python, with experience building apps using Django, Flask, and FastAPI on the backend, and React/Next.js (TypeScript) or Vite for frontends. I've also built basic RAG systems and AI agent frameworks, like this one: [github.com/kliewerdaniel/news17](https://github.com/kliewerdaniel/news17). Here's a blog post walking through the RAG system I built: [danielkliewer.com/blog/2024-12-01-Basic-RAG](https://danielkliewer.com/blog/2024-12-01-Basic-RAG) I’m self-taught via MIT OCW and edX courses, and I enjoy applying higher math to AI problems. One of my projects uses trait-weighted AI personas stored in SQLite/ChromaDB: [github.com/kliewerdaniel/](https://github.com/kliewerdaniel/) — more about that in this book I wrote: [Agentic Knowledge Graphs](https://6340588028610.gumroad.com/l/ddsrtm) That said, I’m a novice when it comes to secure infra and cloud hardening. I’ve dabbled with JWTs and some access control but would need mentorship or time to grow there. My GitHub: [github.com/kliewerdaniel](https://github.com/kliewerdaniel) My site: [danielkliewer.com](https://danielkliewer.com) Happy to chat if you think there might be a fit. --- Let me know if you'd like a version with a slightly different tone (more formal, more casual, etc.).

Sure! Here's a much shorter, clear, and professional reply you can use: --- Hi, I’m interested in the role. I have experience with React, Node.js, Next.js, minimal Express, and REST APIs, mostly using TypeScript on the frontend and Python (Django/Flask/FastAPI) on the backend. I’ve built projects involving AI agents and retrieval-augmented generation (RAG), such as [news17](https://github.com/kliewerdaniel/news17) and wrote about these concepts [here](https://danielkliewer.com/blog/2024-12-01-Basic-RAG). While I’m comfortable with Python and applying math in programming, I’m a novice in secure infrastructure and cloud best practices. You can find my work on GitHub: github.com/kliewerdaniel and my website: danielkliewer.com. Please let me know if you’d like to discuss further. --- Want me to help you tailor it more casual, technical, or formal?