Business Executive to AI Leader
At DataHack Summit 2018 in Bengaluru, a main-stage talk on the data science skills matrix and a working session that turned the same argument into a prioritisation board.

DataHack Summit 2018 ran from 22 to 25 November in Bengaluru, organised by Analytics Vidhya at the NIMHANS Convention Centre. deepkapha.ai was there in two capacities. On the main stage, Tarry Singh gave a talk titled "Climbing that Deep Learning Hill", subtitled a complete overview of your data science skills evolution matrix. Away from the auditorium, in a function room laid out with round tables and flip charts, he ran a workshop for business leaders called Business Executive to AI Leader.
The talk argued a structural point rather than a technical one. Internet companies were absorbing the available talent, leaving enterprises both data poor and talent poor. The deck answered that with a five-plateau model for building skills inside an organisation: fundamentals in maths, statistics and programming, then data visualisation and big data, then machine learning, then deep learning, then applied AI. Each plateau carried a completion figure, from 5 per cent at the first to 100 per cent at the fifth. The deck put the lab's training reach at that point at more than 13,500 learners worldwide.
The workshop turned the same argument into a working session. Groups sat at round tables with printed canvases and worked their own use cases onto a prioritisation matrix pinned to a flip chart, impact of the opportunity along one axis and ease of implementation up the other, sticky notes clustered into quick wins and larger bets. One table's board is labelled "potential killer app". Each group then presented its board back to the room. The slide behind one hand-over reads "Top down: global execution", a project for a multi-country, full-stack transformation.
The workshop closed with certificates of participation, co-signed by Analytics Vidhya and deepkapha, handed over one by one in front of the screen. The cohort photographed together afterwards numbered eleven, a mix of Indian and European participants. Raik Otto presented alongside the deepkapha sessions, on data-driven development of a cancer blood test built on a support vector machine, using differentiation-stage biomarkers.
The archive folder is named 22 to 24 November; every piece of on-site signage and both decks say 22 to 25, so the signage is followed here. The certificate photographed at the handover carries a day digit that will not resolve at any crop, so the summit window is given rather than the workshop day. The 13,500 figure is the lab's own claim on its own slide, and is presented as such. The summit's sponsors on the lobby banner were the summit's, not ours, so they are not listed.








