A lot is being discussed these days on the impact, of Generative AI (Gen-AI), on Indian IT Industry and I see an increasingly gradual formation of consensus view that it’s going to disrupt IT industry (some even saying using the word “extinct”).
A typical argument is, with Gen-AI, the end customers of Indian IT companies will be able to gain significant productivity and cost savings, by driving more automation and reducing the need for IT outsourcing. The other argument, also quite common, is that end customers will be able to in-source a lot of IT work and relegate it to Gen-AI to do it for them, thus reducing the need for IT companies.
I have been associated with the IT industry for last 20 years and although today I don’t work for an Indian IT company I still do a lot of work in Digital space especially around Cloud, AI/ML and automation. And from what I see on the ground, I am not convinced by the doomsday scenario being painted for Indian IT industry.
Below is my take on why Gen-AI is actually going to generate more work for Indian IT companies and be margin accretive than what the consensus view offers.
Let’s understand how the implementation of Gen-AI in businesses will play out and how it will generate demand for more services work than destroy it.
1- To start with, businesses don’t simply spend money on technology just because of hype or its usefulness in public domain (think of Gen-AI writing a poetry/essay). Once businesses decide to spend time and money on Gen-AI, they won’t simply download a software from public domain or assign a bunch of developers to build a software and start using it. From my experience, any truly transformative technology like Gen-AI requires a significant time and cost commitment for organizations to implement, scale, mature and adopt to harness its full potential. The bigger the organization, the more time and cost involved.
2- There are several studies that point out large number of failures of multi-year transformation programs built around new promising technologies. CIOs/CTOs are all too aware of this, many of them having their own share of disappointments with new technologies in the past. So they will be wary of risks. While some of them will wait out the hype cycle of Gen-AI, early adopters will have to work through several aspects such as following but not limited to:
a. Are there any high impact use cases that will generate step change improvement in efficiencies, cost savings or market shares?
b. What will be investment required (think of the costs related to hardware, software, migration, integration, training, support etc) to realize those use cases?
c. Can technology scale up to address complex and evolving organization and operations?
d. What are changes required (think of processes, data governance, IT architecture, skills etc) to sustain the adoption?
e. Are there any potential regulatory, societal, ethical risks from adopting Gen-AI?
3- The bigger the organization and the scale of operations, the more time CIOs/CTOs will spend on analyzing these aspects before pulling trigger on Gen-AI. Most will not have skillset or enough resources to do this in house and they will engage a Consulting partner to help them navigate through complexity and define the right strategy- this will create demand for consulting work.
4- Once strategy is defined, implementing that will require a technology partner for several work-streams associated with Gen-AI adoption such as:
a. Data layer, curation, modeling, enriching etc. Companies can’t simply just use data available in public domain as they will need Gen-AI to work with their own data sets and again from my experience majority of the organizations tend to be quite deficient with their data management.- This alone is expected to generate significant demand for IT services in terms of data work.
b. AI modeling- Again one can’t simply use Gen-AI algorithms available on public domain. They need to work within the context of use case, data and processes of a business- This will create considerable demand for data sciences work to build/tailor the AI algorithms.
c. Hosting- Gen-AI by definition is a highly compute-intensive technology. The typical software that writes essay or poetry is running on massive computers somewhere in Cloud. Many businesses are running on legacy on-prem infrastructure that are not suited to demand of computing resources required to run Gen-AI solutions. They will have no choice but to move to Cloud- This will further accelerate Cloud transition, already underway and is a big part of IT companies’ revenue streams.
d. System integration- Any Gen-AI solution will have to be consumed through some kind of software depending on the use case type and user profiles. Also AI- solution will consume data from businesses’ legacy data systems as well as from public sources- This will bring a lot of work for application development and data integration.
e. Supporting the model- Any kind of AI model (including Gen-AI) will require constant changes as businesses evolve- This will require services for support and maintenance of models and applications.
Depending on the nature and complexity of use cases and legacy systems as well as decisions made for Cloud transition, the skills and efforts required to deliver all this work will vary considerably. Today when businesses are trying to become lean, efficient and focused on their core value propositions, they are not going to start hiring hundreds of developers, architects and data scientists to implement Gen-AI. That’s the reason they have been outsourcing in the first place.
What businesses will do is find the companies who can do it all for them at much lower cost and faster. And to me Indian IT companies will be their perfect choice as the they have already started ramping up their capability and scale to deliver Gen-AI solutions.
People, who doubt if Indian IT companies can move fast on Gen-AI, need to simply look at their past track record with similar highly hailed technologies such as Cloud, Blockchain, IOT, mobility etc. Take Cloud for example: in just a few years, many large Indian IT players (e.g. TCS, Infy) have evolved their capabilities to see regular mention among leaders in research papers published by Gartners’ of the world. I don’t see any reason why they can’t repeat the same success with Gen-AI, thus positioning themselves to win big chunk of work that will come from the adoption of Gen-AI (if and whenever that happens).
Last, I see Indian IT companies have also started to use Gen-AI to transform their internal operations and I believe they will be able to improve their own productivity and margin structure over time with the use of technology.
So, overall I feel rather excited, than unnerved, by prospects of Gen-AI becoming the next big technology wave. Our IT industry has got everything (quality of management, skills, capitals, scale and relationships) to be able to ride this wave than get swept away.
Disclaimer- I have significant holding in IT stocks in my portfolio and can be biased in my views. Pls don’t take it as a recommendation for buy/sell. And as always I welcome different opinions to broaden my perspective.