I'm in the budding stage of my data science learning/career and have been applying to internships for next summer as a first year at community college. With over 100 applications, and 4 interviews total, I was just rejecting from my closest chance yet, so I have decided that I need to create my own internship. My thing has always been that I need to get my hands on real-life data, particularly data that involves transactions or money, and since I live in a small town, offering to analyze local restaurants' data seemed like the perfect opportunity to learn and potentially make some money in the future.
I live in a town with ~30,000 people. Suffice to say, there are a lot of small/medium-sized and mom-and-pops restaurants/shops in town. We have a bustling downtown scene, so there are a lot more restaurants and boutiques than one would normally think for a town of our size.
The premise that I have so far is that I would approach 1 or 2 mom-and-pops restaurants (would start with the local pizza place that I live a couple blocks from) and offer to analyze their data for general insights/trends that they could capitalize on to make more informed decisions. I would offer the first one or two consultations for free, and if I recognize that I'm knowledgeable/confident enough and they think what I do can benefit them, then we would ideally start working contractually.
I am aware that this would qualify more as DA/BI, because the extent of my ML knowledge is classification and regression. I imagine that time-series forecasting would be the modelling most relevant to restaurants, and is something that I may look into in the future.
So far, in terms of my value proposition, I have:
- Dish popularity/menu analysis
- Price adjustments based on trends
- Customer demographic analysis
- Inventory management/food waste analysis
- Discount/loyalty analysis
- Seasonality trends/insights
- Payroll/staffing analysis
- Delivery analysis (since they do pizza delivery)
I won't dive deeper into each of the points here, but what I would like from this post is:
- Is this a good idea in the first place
- Are there any red flags that you see in this idea/process
- What can I do to build upon this idea
- To what extent is this plan anything other than just building dashboards
Subreddit
Post Details
- Posted
- 2 years ago
- Reddit URL
- View post on reddit.com
- External URL
- reddit.com/r/datascience...