Building B2B SaaS ML Features Overview
2 min readJan 31, 2022
Compared to B2C Space, creating Machine Learning features in B2B Space is more complex and requires a long cycle of engagement.
In this series of articles, we will go through step by step in understanding and building ML Feature in B2B Space
- Difference Between B2C and B2B: This article understands the difference between and B2B and B2C. We study four real-world cases studies.
* Canned Response Suggester (CRS)
* Field Classification in Customer Support (FC)
* Customer Churn Prediction (CCP):
* Sentiment Analysis in Social Media Brand Reputation (SA): - Before Model Building: This article talks about the things that need to be taken care before model building, such as collecting right data, data exploration, refining metrics and getting Product Management on the same page.
- Model Building: This article talks about the typical model building process, i.e. how to choose right feature and right models and how designing the ML feature is different to Kaggle competition.
- Model Building in B2B: This article talks about the ML model building techniques specific to B2B space.
- Model Improvement in Production: This article talks about the steps that one should follow for monitoring and continues improvement of the model
- Deploying ML Pipelines for B2B: This article talks about the techniques can deploy ML pipelines for quicker and better deployment.
- Choosing Machine Learning Career Path: This article talks about the job opportunities that are available, how ML is developing and what companies are best suited.
P.S Thanks for reading. My experience is more in the startup space. Thus, this list of articles will miss out on some of the best practise. It would be great if you could point it out. We can discuss it over tea, pizza or whisky or whichever you are comfortable with.
Token of Gratitude:
I am grateful that the list of peoples who were patience to my complaints and gave me insights that helped me fill my gaps
