Generative AI has been a disruptive force and the Indian IT sector has been quick to recognise its potential. Leading IT companies in India are swiftly harnessing the power of generative AI and delivering its advantages to their clients. In April, Tech Mahindra became the first IT giant to launch something like a Generative AI Studio. Tech biggies, including Tech Mahindra, are making substantial investments in training their workforce to become proficient in this transformative technology.
“Generative AI is bringing a renaissance in the creator economy. Until now, AI has demonstrated its capabilities in observing, detecting, recommending, and predicting. Generative AI can create as well, thus making AI further disruptive. Looking at its tremendous potential, we are focused on investing in, and implementing this technology effectively,” Hasit Trivedi, CTO, digital technologies, and global head – AI at Tech Mahindra, told AIM
In an exclusive interaction, he revealed that the IT giant was one of the early adopters of generative AI. Way before the launch of ChatGPT, the company developed the Storicool platform, an auto content creation tool, which was beyond its years, according to chief executive CP Gurnani. Tech Mahindra is also building an Indic large language model (LLM), something which could be seen as a competitor to OpenAI’s GPT models that power the popular chatbot, ChatGPT.
Generative AI Studio
Earlier this year, Tech Mahindra launched its Generative AI Studio under its amplifAI0->∞ suite of AI offerings and solutions. “We were the first to launch a generative AI studio for experimentation and also use it internally,” Trivedi said. The studio is a one-stop centre to experiment, generate and optimise digital content. “It will help enterprises make the right decisions, identify the right technology in the space of generative AI, and aid them in achieving the next level of digitised efficiency. The studio enables enterprises to experiment with different tools, do the first level of proof of technology, and also run a small pilot with the selected technology for further experimentation.”
The studio will consolidate six aspects of content generation — code, document/text, image, video, audio, and data — under one umbrella, empowering enterprises to discover and explore the wide-ranging possibilities of generative AI. “The Generative AI Studio also provides enterprises with a user-friendly interface and a range of features to customise their content. Users can select their preferred content type, customise their options, and then let the studio handle the rest of the process.”
Moreover, the studio has codified prompt engineering in a form that reduces randomness in responses, and the standardisation approach was adopted to generate content, taking advantage of NLP – which is extremely beneficial for enterprises.”
Being the first to launch a generative AI studio, Tech Mahindra is also leveraging the technology internally for application development, testing, migration, legacy modernisation, transition, documentation and so on. “As a part of BPS business, we are using it in service desk, customer servicing, reporting and reconciliation, etc. In our internal functions, we are using it in legal, CIO, HR, and marketing functions. In an AI lifecycle, we are using it for synthetic data generation and annotation.”
Furthermore, Trivedi revealed that Tech Mahindra is effectively adopting generative AI in its mainstream AI projects through Generative AI studio, TechM XaaS (Xperiment as a service), Evangelise Pair Programming, Enterprise Knowledge Search, and other cutting-edge solutions that are powered by generative AI. “We are infusing it into most of our technology services, including cyber security, metaverse, business excellence, intelligent automation, and advanced analytics elements to increase efficiency and productivity.”
Bringing gen AI capabilities to customers
Currently, Tech Mahindra is already implementing generative AI use cases for many customers, notably a leading multinational oil & gas transport pipeline company. “We have automated the generation of transcription of each discharge summary for a healthcare player, thereby helping to reduce efforts and turnaround time from days to a few hours.”
Moreover, Tech Mahindra’s generative AI offering is helping a global technology distributor in providing reports using machine learning techniques like NLP and NLG / NLU to generate text. “The solution ingests raw data to generate text for the outcome analysis report, and it provides reports which have analysis, observation, analytics with visualisation, business insight, and recommended actions.”
As a result, they have seen a drastic reduction in the time taken to process documents, curb costs, reduce dependency on domain experts, and curate a consolidated view of the data extracted. Furthermore, Trivedi revealed that the IT giant is also keeping an eye on acquiring products that enable them to build solutions involving advanced generative AI capabilities. “We will continue to foster new partnerships, especially with startups or products with niche capabilities that enable us to provide the right solution to our clients.”
Ethical and responsible use of gen AI
In light of the extensive uptake of generative AI by enterprises, it is crucial to conduct a thorough evaluation of the associated risks. These include concerns related to cybersecurity, data protection, copyright issues, as well as ethical and legal implications. Besides, till date, hallucination still remains a problem with LLMs. However, Trivedi believes hallucinations are reasonably manageable using techniques available, especially when dealing with enterprise data.
“Our Responsible AI framework allows clients to address these issues through multiple mechanics like process, tooling, techniques, awareness, education, and involvement of cross-functional teams. There are platform/product specific capabilities which need to be leveraged in the right manner. At times there are limitations in the said product/platform, which first need to be known and then addressed by appropriate mechanics,” Trivedi said.
Moreover, as part of the “Responsible AI” drive, Tech Mahindra has instituted many actions, like ensuring that the employees are trained and aware of the responsible AI aspect, putting the right governance framework to govern AI projects, among other things. “We also encourage human-in-loop approach in critical AI-led inferences and actions and always ensure a structured, comprehensive and responsible AI maturity assessment for clients. Thus, enterprises across industries can leverage Tech Mahindra’s comprehensive and customisable on-premises and cloud-based solutions with the assurance of harnessing the remarkable potential of generative AI and driving impactful business outcomes.”
