RAG exercise completed

README.md

@@ -1,9 +1,2 @@
-# Gradient Works Interview Repo
-
-Hello friends! If you're here, you're probably looking for an engineering job at Gradient Works and I'm excited you're here!
-
-This repo is a collection of exercises that we use to evaluate candidates for engineering roles at Gradient Works. We hope you find them interesting and challenging. We also hope you find them a good representation of the kind of work you'd be doing here.
-
-If you're looking for the Machine learning role, please check out the [machine learning exercises](./machine-learning/README.md).
-
-We're not currently looking to hire someone for a Salesforce engineering role, but if you have a morbid curiosity around that, you can check out the [Salesforce exercises](./salesforce/README.md).
+## RAG based LLM 
+### model: GPT 3.5

machine-learning/README.md

@@ -1,46 +1 @@
-# Gradient Works Machine Learning Interview Repo
 
-Hello friends! If you're reading this, you're probably interviewing for a machine learning position at Gradient Works. If not, GET OUT! SHOO! Go on now, git! Just kidding, you can hang out. But you're probably not going to get much out of this repo. Maybe you will, who knows? I'm not your mom.
-
-Anyway, if you're here for the interview, welcome! I'll outline the steps for the whole process here and if you have any questions, send an email to mando@gradient.works and he'll clear things up.
-
-## Step 1: Phone screen
-
-You've probably already done this, but if not, we'll schedule a 30-minute phone call so we can get to know each other. We'll chat about your background and experience, and we'll also talk about Gradient Works, the role and what you're looking for in your next job. If everything is a good fit, we'll move on to the next step.
-
-## Step 2: Technical exercise
-
-We'll send you a technical exercise to complete. It's a Jupyter notebook that you can run in your own environment or in Google Colab, plus a tiny CSV dataset. The exercise will test your ability to explore data, answer questions about that data, and build a small RAG system to answer arbitrary questions about the data.
-
-We expect the exercise to take 1-3 hours, depending on how far you take the RAG system. We're not looking for a perfect solution, but we are looking for a thoughtful approach and a good understanding of the approach (there's more detail in the notebook).
-
-You can find the exercise [here](./exercise.ipynb).
-
-To complete the exercise, you can either:
-
-1. fork the repo and submit a pull request with your solution, or
-2. import the notebook into your own Google Colab environment and share the notebook with mando@gradient.works
-
-If neither of those work, email mando@gradient.works and we'll figure something out.
-
-> [!NOTE]
->
-> If possible, we'd like to invite you into our Slack while you're working on the exercise. This way, you can ask questions and we can get to know each other a bit better. If you're interested, let us know and we'll send you an invite.
-
-> [!NOTE]
->
-> You might be asking yourself, "Doesn't this dummy know that I can just look at the notebook and get started early?". And yeah, I get it. But the whole point of this is for you to demonstrate your abilities and level of understanding. So, if you spend a bunch of time learning all the ins and outs of RAG systems ahead of time, well, I mean that's great for us :).
-
-Once you've completed the exercise, we'll review it and schedule a Zoom interview to discuss your solution.
-
-## Step 3: Zoom interview
-
-We'll schedule a 1.5-hour Zoom interview to discuss your solution to the technical exercise. We'll ask you to walk us through your approach, explain your code, how you might extend the solution, etc. We'll also make sure to leave plenty of time for you to ask us questions about the team, the company, and what we're building and how.
-
-After that, we'll make a decision and let you know as soon as possible.
-
-## Step 4: There's no Step 4
-
-Hopefully this all makes sense and you're excited to get started. Interviewing is hard and we get that. We hope that this process makes sense and is as stress-free as it can be.
-
-As always, feedback is a gift, so if you have any thoughts on how we can improve this process, please let us know.