Technology | MSME | Build | Future | Trends

Notes on Indian IT SMEs

recent posts

about

  • Kochi becoming global traffic hub

    Kochi becoming global traffic hub

    Is it possible for Kochi to become upcoming global traffic hub? Its geographic location and direct connection to undersea cables certainly make it a strong candidate. While there are natural challenges such as heavy rainfall, high humidity, flooding, and coastal corrosion, these risks can be managed through proper disaster mitigation strategies. If addressed effectively, Kochi could leverage its strategic position to support faster, more resilient data centers. Drawing lessons from established hubs like Singapore and California, the city could position itself as a global gateway.

    Building high‑speed network infrastructure here would also benefit nearby states, including Bengaluru, by offering alternative connectivity routes. Additionally, complementary locations such as Tamil Nadu could serve as backup or distributed sites, similar to how Colorado and Utah complement other U.S. hubs, to minimize risks from natural disasters.

    With land availability and rising costs pushing data centers toward Tier‑2 cities, Kochi stands out. The city offers a skilled workforce, a growing IT ecosystem, increasing investments in renewable energy, and strong potential for expansion. While attracting talent from other regions remains a challenge, the trend of GCCs (Global Capability Centers) shifting to Tier‑2 hubs, and major companies planning offices in Kochi, will bring more professionals into the city.

    To realize this vision, the most critical requirement is reducing bureaucratic hurdles—such as delays in approvals and infrastructure clearances—and enabling faster execution of projects. Beyond meeting local enterprise needs, Kochi has the opportunity to establish itself as a major digital gateway, complementing India’s Tier‑1 cities and strengthening the country’s global digital infrastructure.

  • When Truth Fragments

    When Truth Fragments

    Recent news about political bias in chatbots highlights how far we’re drifting from truth. Historically, information—often half-truths and fiction—was used to impose order, not just reveal truth. Over time, these half-truths became accepted wisdom. As communication evolved from religious manuscripts to the internet, such ‘educated fictions’ multiplied, fragmenting society with competing narratives. What was once limited to religious conflict now divides us by politics, region, language, and more. Today’s influencers, more educated than ever, create fact-based fictions that shape people’s beliefs more deeply than simple traditional stories.

    Search and social apps already create biased echo chambers, and new generations of applications may become even more extreme. Major companies like Google and Meta have yet to solve this personalization problem. As more foundational models become widely available, half-truths risk being further distorted without proper safeguards. Many users, unable to distinguish fact from fiction, are content as long as they hear what they want—often realising the consequences only after significant harm is done.

    For example, YouTube addiction among India’s elderly is driving extreme thinking and resistance to other views. This generation, once informed by newspapers, now places similar trust in new media, leading to harmful outcomes. As chatbots become popular, their influence on younger generations could result in collective failure across all ages.

    Solutions require platforms to take responsibility for their content, with transparent and diverse review panels to reduce bias. Users should have clear controls over what they see, not just rely on platform owners. Robust, ethical regulations—set independently by both governments and companies—are essential safeguards.

    In the end, we must strive to foster a more informed, connected, and humane society.

    References

  • AI Adoption in organisation

    AI Adoption in organisation

    AI adoption should be approached from the top down, with C-suite oversight playing a critical role. Small and medium-sized enterprises (SMEs) often lack the budget for extensive experiments or ready-to-ship products. This process involves significant change management and costly decisions, which can lead to challenges at key decision points. Updating business processes requires multiple iterations to ensure that they are suitable for the organization, necessitating a careful balance of motivation, time, and financial resources.

    To drive motivation or ROI focused organizations, they can pursue specific transformation goals. For example, they might aim to increase revenue by 30% through AI, boost productivity by 50%, or introduce new business models. For companies willing to spend money, investing in foundational infrastructure and all-in-one solutions is often more beneficial than a step-by-step approach, as it provides greater clarity and reduces the need for rework across interconnected systems.

    Transformative thinking is essential for executives, who must continuously evaluate and enhance processes through machine collaboration. Instead of investing in expensive training programs, SMEs can find and recognise AI Champions to promote self-learning and facilitate the overall goal of achieving AI knowledge within the organisation.

    Establishing a non-revenue-driven Center of Excellence (CoE) can be advantageous for many organizations. For companies with limited budgets, forming smaller, regular teams can facilitate steady progress. The CoE should operate cross-functionally, coordinating with various departments to identify problems and solutions, document improvements, and assess the return on investment for each initiative.

    Cultural attitudes significantly impact AI adoption, particularly since failure is often expected in new domains. Organisations must cultivate a culture that embraces learning from failures, as this mindset is essential for navigating challenges and ultimately achieving success.

    AI initiatives fundamentally focus on collaboration with machines, allowing executives to determine the extent of machine involvement and design effective inputs and outputs. Based on the level of collaboration desired, organisations can also offer specialised upskilling to ensure that employees are equipped to work effectively alongside AI technologies.

    Although AI governance has been less emphasized in recent years, it is becoming increasingly critical for many organizations. SMEs may find themselves relying more on third-party consultants rather than building internal governance teams to manage AI-related challenges effectively.

    Successful organisations employ strategic approaches that encompass clear strategic roadmaps, a focus on organisational capabilities, and an emphasis on effective change management to ensure the successful adoption and scaling of AI initiatives.

    References:

  • Indian IT Sector & US Tariffs

    Although the Indian IT sector is not directly impacted by US tariffs, the indirect effects could be significant. With 55-70% of revenue coming from the US market, the tariffs are likely to have visible secondary or indirect impacts in the coming quarters. Concerns and uncertainties in business processes and services are already prevalent, as companies strive to deliver more efficient, value-added solutions through AI. Additionally, growing fears of U.S. inflation and recession are further complicating the landscape, making it increasingly difficult to initiate new ventures and projects.

    Companies manufacturing laptops, servers, network equipment, IoT devices, and other IT hardware for export to the US may be directly affected. These uncertainties could prompt investors and customers to renegotiate contracts or add clauses addressing these geopolitical risks. Planned expansions or projects might also face increased scrutiny from investors due to these uncertainties and risks.

    Diversification

    One solution to mitigate these challenges is diversification:

    • Market Diversification: Leverage opportunities in other markets beyond the US.
    • Product Diversification: Expand into other non-tariffed products or categories.
    • Supply Chain Diversification: Capitalise on domestic consumption of raw materials and promote local products to boost the economy.

    Higher-margin, value-added services

    Another potential area is higher-margin, value-added services. Establishing specialized, deep centers of excellence by combining domain expertise with AI research can create competitive advantages. India could prioritize domains where it already has an edge, offering solutions with intellectual property rights and ownership. Since current AI models and algorithms are largely based on training datasets from Western countries, there is a significant opportunity for India, being a data-strong nation, to redefine these models.

    Revenue Rethinking

    Furthermore, companies can consider shifting their revenue models. Moving away from traditional Time & Money billing to models based on a percentage of revenue, cost savings, or outcomes could generate continuous income as services are consumed. Additionally, offering services in an as-a-service model can foster stickier ecosystems with recurring revenue streams.

    Overall, while US tariffs may not directly affect the Indian IT sector, the secondary impacts require strategic adaptation through diversification, investment in higher-margin services, and innovative revenue models to sustain growth in an uncertain global marketplace.

  • AI Toolkits for small scale enterprises

    In today’s fast-paced digital landscape, new tools are released almost daily, making it increasingly challenging to even try them out once. Often, we find ourselves reading about these tools instead of using them. We tend to stick with the first tool we adopted and avoid the effort of learning something new. Even if we do switch to a new tool, there’s always another one emerging the next day with even more exciting features. Despite the market competition and the overwhelming number of tools available, those who use at least one of them know they can significantly free up valuable time. This immense power is something Atlas Shepherd experienced when working with Smith (ARC).

    Here are some cost-effective AI tools that can boost productivity and operations for small-scale IT enterprises:

    UI Designs

    • Canva: Free version allows 5GB cloud space, and single user. Mainly for create professional designs.
    • Adobe Sensei: This should be a value add for anyone who already in the Adobe world.

    General Code Assistance

    • GitHub Copilot: for general coding support, works well with VSCode. 
    • Codium: For general coding support.
    • AWS CodeWhisperer: Tailored for AWS-specific tasks.
    • V0 by Vercel: For web development, mainly JS & Python.

    Rapid Prototyping

    • Replit Ghostwriter: A cloud-based tool for fast iteration and quick prototyping.
    • bolt[.]new and lovable[.]dev: Excellent for quickly creating prototype web applications.
    • cursor[.]ai: Another robust option for rapid development.

    Local Deployment Solutions

    • Tabnine: Offers privacy and security along with intelligent coding assistance.
    • Llammastack & client SDK: This can be used to create customisable llama apps.
    • CodeT5: Code understanding and generation, open source code from Salesforce. Customisable, not maintained, and effort should be required.

    Quality Assurance Tools

    • Katalon: Ideal for API and UI testing, enhancing overall efficiency.
    • LambdaTest: Useful for ensuring cross-browser compatibility.
    • Functionise: Customisable for enterprise solutions.
    • Test[.]ai: Automates mobile testing processes.

    These tools can help manage 1-2 small projects, with the limited free tier. Many other tools offer trial periods ranging from 14 to 30 days, which can be impractical unless conducting a case study for an existing larger account.

  • How smaller companies can adopt AI

    Everyday reading news about Large companies are investing millions for AI. So we all end up in the innovators dilemma. Rapid pace at which innovations and solutions are poping up is worrying us. Best in market solutions are iterating many times, rethinking future user interactions. AI solutions, and Agentic AI market is getting overcrowded with thousands of alternatives like SaaS market now. 

    Smaller companies lack infrastructure, training resources, dedicated R&D teams to the help catch up the race. Even when they are ready to start, how/where to start, with affordable and beneficial and not just for copying others.

    How

    Four A’s, (i) Awareness, (ii) Assessment, (iii) Architecture and (iv) Acceleration, suggestion seems like a feasible structured framework for organisations to adopt(link). 

    Where

    If we are about to adopt AI in the organisation, where should we implement? We could optimise existing operations like HR, Finance, Sales, etc, any administrative repetitive tasks. Some of those operations like onboarding new employees, enrich personal info of a potential candidate, researching about competitors, cold email sequence, better marketing contents, etc. Thereby improving productivity, efficiency, and reduce costs.  

    Who

    Change management is hard. Find the right set of motivated people and form an internal team, who should be leading the change, thereby motivating others. If your industry is IT services, we could do a study on an easy going project, what are the specific internal challenges. Once we can find the positives, do same on a bigger project/team. 

    This gradual approach can spread the positives through out every team.

  • Sharing a Framework

    Overview

    Everyone wants to share some piece of reusable code, or a set of functionality with others. We will take a look at a generic and simpler way of creating and sharing a Framework in Swift.

    Support

    It should support main packaging solutions like Swift Package Manager and Cocoapod. We will be keeping the structure in accordance with SPM.

    Steps

    1. Create a Framework project.
    2. Once created, open the root directory and create a folder named ‘Sources’ and another called ‘Tests’. Move the folder named the target(in this case ‘Framework’), with an umbrella header file inside the Sources.
    3. Remove the existing ‘Framework’ folder reference in project, and add the ‘Sources’ and ‘Tests’ folders with ‘create folder reference’ into the project.
    4. Add your reusable source codes inside the ‘Sources’ with correct access control levels.

    Now we have a framework project, which has some reusable component. Next step will be to add support for Swift Package Manager and Cocoapod.

    Swift Package Manager

    For the framework to be available in Swift Package Manager, we need to add a Package manifest file(Package.swift). This will have information about the Framework, its dependencies, targets, etc.

    To achieve it, run swift package init, from project root directory. It will create manifest, README and gitignore files.

    ~/Documents/projects/Framework 6:11$ swift package init
Creating library package: Framework
Creating Package.swift
Creating README.md
Creating .gitignore

    Run swift build, to make sure the project is runnable.

    If in case, you didn’t create unit test with project creation, it will show error saying:

    error: 'framework': Source files for target FrameworkTests should be located under 'Tests/FrameworkTests', or a custom sources path can be set with the 'path' property in Package.swift

    For this, go inside the framework project and just add a unit test bundle target. If created from project, it will be created inside the root directory, then move to ‘Tests’ folder.

    Now run swift build, and make sure build is complete.

    ~/Documents/projects/Framework 6:19$ swift build
Building for debugging...
[2/2] Compiling Framework Reusable.swift
Build complete! (0.56s)

    Sharing SPM

    For sharing the Swift Package Manager, we will be tagging the git version with the manifest file and push the tag to origin. For development purpose, drag and drop the package to the Xcode project/workspace(editing local package dependency)

    Cocoapod

    For initiating the a podspec file, run pod spec create <FrameworkName>. Update the podspec with relevant information.

    Pod::Spec.new do |spec|

  spec.name         = "Framework"
  spec.version      = "0.0.1"
  spec.summary      = "A short description of Framework."
  spec.description  = <<-DESC This is a sample description for the Cococpod library. 
                   DESC

  spec.homepage     = "http://github.com/bob/Framework"
  spec.license      = "MIT"
  spec.author             = { "bob" => "bob@email.com" }
  spec.platform     = :ios
  spec.platform     = :ios, "15.0"

  spec.source       = { :git => "http://github/bob/Framework.git", :tag => "#{spec.version}" }


  spec.source_files  = "Sources/Framework/*.swift"
  spec.resources = "Sources/Framework/Resources/*.xcassets"

  spec.dependency "JSONKit", "~> 1.4"

end

    Once done, use another project and point to the local podspec and update the pod(developing pod).