Notes from eChai's DeepTech Startups Meetup at Ahmedabad University

How drivebuddyAi started and what drivebuddyAI does?

DBAI is an Automotive AI startup working in the domain of Automotive Aftermarket in the category of ADAS.

What is AI in drivebuddyAI and what are the challenges about Human AI Interactions?
The drivers are usually in touch with our driver trainer team who talks to drivers and comforts them for their better safety and behavioral issues identified by the technology.

How do drivers react to AI alerts and how the resistance is being handled?

The drivers take 10-12 months of efforts to get confidence and comfort with the AI alerts i.e. a typical human psychological phase of accepting a tech in daily work life.

⁠What are the challenges of getting a customer, investors?

DBAI is funded and was raised before Covid in 2020 and since then have progressed towards becoming one of the major players in Indian market.

What are the challenges about hiring talent, building AI models and how it was handled in the case of drivebuddyAI?

Ahmedabad specifically lacks in talent pool required for AI, dbAI has talent pool available in Delhi, Vizag, Pune, Jaipur and all these are mid experience to entry level engineers working under the managent of domain expert AI engineers to solve daily AI challenges.

Typical challenges of building AI involves data collection, data curation, annotations, training, GPUs, model optimization etc. 

The DBAI journey started at a very early stage by using the pre-trained models for perception module to collecting and annotating data for Indian ecosystem when there was no Indian dataset available for building AI on it. The process of annotations, having GPUs for training purpose and managing a complete cycle of the data generation to updating a model with continuious learning is the journey in building AI.

To build AI/deep tech product startups, first one needs to start with whatever is readily available in the market which can give functional accuracy and then building on top of it by the feedback from the users, customers can help building the product focused on a specific domain.