Bangladesh Local players eye bigger slice as in-country Cloud Computing storage/processing gains foothold

[-blitzkrieg-]

Obviously you have no idea about local e-Governance, and Fintech initiatives in Bangladesh - hence your negative comments.

Low potential ?

In a country of 170 Million people ? Which is half the population of the US ?

Bangladesh - despite its size has the second largest pool of freelancers globally and actually exceeds India in some sectors.

Coding is not the only game in town when it comes to online access and services business.

And we don't have to beg AWS, Azure and GCP to open datacenters in Bangladesh, when we can spend money and build our own.

Dude where did you pick that chart from ..atleast reference it. Because that's not how top international freelancers look like

The 10 countries with the fastest-growing earnings for freelancers

Technology has made it easier than ever for individuals to boost their earnings with profitable side-gigs. And while U.S. workers of all ages are reaping the greatest rewards, young freelancers across Asia are quickly getting in on the act.

www.cnbc.com

 

[leonblack08]

Dude where did you pick that chart from ..atleast reference it. Because that's not how top international freelancers look like

The 10 countries with the fastest-growing earnings for freelancers

Technology has made it easier than ever for individuals to boost their earnings with profitable side-gigs. And while U.S. workers of all ages are reaping the greatest rewards, young freelancers across Asia are quickly getting in on the act.

www.cnbc.com

Your reference relates to revenue growth. His chart refers to market share. Both are not comparable.

But yes, would like to see the source of your chart @Bilal9 bhai.

 

[Bilal9]

Dude where did you pick that chart from ..atleast reference it. Because that's not how top international freelancers look like

The 10 countries with the fastest-growing earnings for freelancers

Technology has made it easier than ever for individuals to boost their earnings with profitable side-gigs. And while U.S. workers of all ages are reaping the greatest rewards, young freelancers across Asia are quickly getting in on the act.

www.cnbc.com

Most people in the West are completely unaware of how many freelancers exist in Bangladesh, this writer for CNBC is a prime example.

Freelancers from Bangladesh produce a lot more complicated and higher value addition work than garbage tech support work (or its illegal off-shoot - scamming seniors in the West) which India does. I personally know dozens of people who do CAD, CGI, desktop illustration, ad design, tax and medical transcription work. There are half a million freelancers working in Bangladesh. That said, freelancers working in Pakistan may be just as many if not more.

Here is the article acknowledging the same - it was done by Oxford Internet Institute, which is a subsidiary of The University of Oxford.

OII | Where are online workers located? The international division of digital gig work

www.oii.ox.ac.uk

Government Initiatives to Develop the Freelancing Sector – Freelancers Bangladesh

freelancers.gov.bd

Oxford: Bangladesh second largest source of online workers

Bangladesh provides 16% of online labourers, according to a study from Oxford

archive.dhakatribune.com

 

[Bilal9]

Have to be honest do not know BD scenario at the coalface as never worked there but i am assuming BD will soon reach a point where companies will seek to exploit market data. I am thinking the likes of Uber alternates, mobile banking platforms, amazon type startups etc

Exploitation of big data is going to be big business very soon in BD. Likes of MS can be made to invest only if BD geofences its data now.

Otherwise it really is not rocket science to put an alternative service together. BD already have the basic infrastructure.

My thoughts exactly. That is why we need a mini-GDPR type data-privacy regulation/mandate for Bangladesh and apply it to geofencing our data like you said.

The local govt. "experts" only care however when a data regulation can be used to legally frame and persecute the opposition, it's time they took a longer view and saw where the future prosperity of our country is at stake.

I am sure if they went abroad to UK and Silicon Valley, they could find tons of Bangladeshi experts who can help them do exactly this.

I may be crying in the wilderness however. These govt. idiots will most probably look toward some overpriced "Dada Consultancy Firm" from Kolkata or Delhi to come in and help them out. Helping local Bangladeshi people or companies is against their nature. No wonder expat Bangladeshis never come back to their home country once they leave.

 

[Pranaayam]

Data centres exists in BD as I have already said. How much effort do you think is involved in repointing or in that matter spinning up clusters to process data within BD setup - nothing at all.

To answer your first point..... yes no software change is necessary. You have to configure your solution to say where your data is, pointing to a DC in BD is no greater effort than in pointing to US. Additionally you get charged for data movement, everytime BD data leaves or re-enters BD it will be charged. Furthermore given where the ultimate destination is and how many global zones the data crosses has bearing on storage and compute costs.

Just lead a project to hive off my companies Russia business from the rest of the world. In 2 months we moved off Azure and AWS to Yandex with approximately 3PB of data. Russia never allowed its data to leave its jurisdiction, its data stayed in its own data centres. Switching cloud providers is easy if you control the data.

Why should BD care for what is easy for MNC? As i said they do not need to come to BD if they do not want to. There are many players in the market who are nimble enough to accommodate jurisdictional ring fencing.

Ok, then. Since it makes economic sense to you, I hope Bangladesh enforce data localisation.

Most people in the West are completely unaware of how many freelancers exist in Bangladesh, this writer for CNBC is a prime example.

Freelancers from Bangladesh produce a lot more complicated and higher value addition work than garbage tech support work (or its illegal off-shoot - scamming seniors in the West) which India does. I personally know dozens of people who do CAD, CGI, desktop illustration, ad design, tax and medical transcription work. There are half a million freelancers working in Bangladesh. That said, freelancers working in Pakistan may be just as many if not more.

Here is the article acknowledging the same - it was done by Oxford Internet Institute, which is a subsidiary of The University of Oxford.

OII | Where are online workers located? The international division of digital gig work

www.oii.ox.ac.uk

Government Initiatives to Develop the Freelancing Sector – Freelancers Bangladesh

freelancers.gov.bd

Oxford: Bangladesh second largest source of online workers

Bangladesh provides 16% of online labourers, according to a study from Oxford

archive.dhakatribune.com

I can't but laugh at your stupidity. You have zero understanding of Indian IT ecosystem.

 

[mb444]

Ok, then. Since it makes economic sense to you, I hope Bangladesh enforce data localisation.

I can't but laugh at your stupidity. You have zero understanding of Indian IT ecosystem.

Data is a resource just like oil for example and must be seen as such and protected as such.

I am not BD based so have little direct knowledge about the specific condition. However it goes to reason that a nation of 170m generates a lot of data and can be harvested and used to direct policies/targeted support/initiatives/legal compliance by government and investment/business descisions/process improvement by the private sector.

This will all come to be as it is already happening in advanced economies. BD needs to take steps with future in mind that protects its data and incubates local solutions.

 

[Pranaayam]

Data is a resource just like oil for example and must be seen as such and protected as such.

I am not BD based so have little direct knowledge about the specific condition. However it goes to reason that a nation of 170m generates a lot of data and can be harvested and used to direct policies/targeted support/initiatives/legal compliance by government and investment/business descisions/process improvement by the private sector.

This will all come to be as it is already happening in advanced economies. BD needs to take steps with future in mind that protects its data and incubates local solutions.

You see the cloud is not just data. You need to process the data (requires huge compute power) to generate meaningful inferences. This is where major cloud providers will win hands down with all their in-house data processing and ML services. But you seem to think that just by providing storage services and with data localisation law, you can push big tech to change their stand. Not going to happen. They don't see revenue potential in Bangladesh yet. Sorry for being direct.

You can have local apps use the data storage of the country but even they would want use big cloud providers given the reliability (because of SLA) that comes with it, unless you think Bangladeshis can compete with big cloud providers.

Max to max some government apps can be mandated to use local data services and no one else. It just doesn't make economic sense.

I hope Bangladeshis law makers are smarter than you.

Even India uses AWS for its government related apps, lucky for India we have AWS data centers in India.

With CoWIN, India has prioritized in making platform scalable: AWS health lead Nicola Murphy

Murphy spoke to BT about how the technology can benefit India and the world beyond COVID-19 pandemic.

www.google.com

 

[mb444]

You see the cloud is not just data. You need to process the data (requires huge compute power) to generate meaningful inferences. This is where major cloud providers will win hands down with all their in-house data processing and ML services. But you seem to think that just by providing storage services and with data localisation law, you can push big tech to change their stand. Not going to happen. They don't see revenue potential in Bangladesh yet. Sorry for being direct.

You can have local apps use the data storage of the country but even they would want use big cloud providers given the reliability (because of SLA) that comes with it, unless you think Bangladeshis can compete with big cloud providers.

Max to max some government apps can be mandated to use local data services and no one else. It just doesn't make economic sense.

I hope Bangladeshis law makers are smarter than you.

No. Every SLA that likes of Azure have revolves around data redundancy or how many copies of the data one holds. Storage is first priority.

Then you move onto processing data. Here you have to build your own data processes using databricks or whatever you want. There are slas around uptimes of cluster and how elastic the resource provisioning will be.


You are mistaking cloud with an ERP system the like of sap or oracle that comes with best of breed data models and processes. No such things exist in cloud.

There is nothing BD need do that many countries have not already done. Big providers are not necessary when it comes down to it. Its fine to have your view but is not something that many policy makers would share. You serve the interest of your country not foreign IT companies.

 

[sonargon]

You can have local apps use the data storage of the country but even they would want use big cloud providers given the reliability (because of SLA) that comes with it, unless you think Bangladeshis can compete with big cloud providers.

You don't understand. At the end of the day it doesn't matter if big cloud providers provide objectively better service. There is no big culture of cloud in BD yet. Use of certain software tools/infrastructure can be cultural - just look at Japan or China (AliCloud). So rather than let the growing BD IT professionals fall into using the international cloud providers - we can instead use this moment to re-direct them to use local Cloud services. Of course the onus is on the local cloud services to provide outreach and training on their offerings - and there should be a strong culture of feedback so they can tailor their services and support specifically to the needs and workflows of Bangladeshi IT professionals (mostly web-devs).

The big cloud providers can handle "advanced services" - aka AI/ML workloads. 90% of apps don't need such capabilities Let the local cloud providers focus on the subset of "solved" infrastructure (VM, Storage, DNS, CDN, DB Hosting). And of course there is nothing stopping a BD startup from storing their data on a local cloud provider and then transferring subsets of this data to a big cloud provider for AI/ML processing on an as-needed basis.

This is better for the country in the long run. It lets the people develop cloud services from the ground up and truly understand how it all works from the fundamental building blocks - rather than skip this process and go to a big provider who has solved all the challenges. You should not be proud that India has opted to skip this in favor of AWS. It robs your country-men of a truly valuable learning experience that would only enhance their overall technical competency.

 

[Bilal9]

You don't understand. At the end of the day it doesn't matter if big cloud providers provide objectively better service. There is no big culture of cloud in BD yet. Use of certain software tools/infrastructure can be cultural - just look at Japan or China (AliCloud). So rather than let the growing BD IT professionals fall into using the international cloud providers - we can instead use this moment to re-direct them to use local Cloud services. Of course the onus is on the local cloud services to provide outreach and training on their offerings - and there should be a strong culture of feedback so they can tailor their services and support specifically to the needs and workflows of Bangladeshi IT professionals (mostly web-devs).

The big cloud providers can handle "advanced services" - aka AI/ML workloads. 90% of apps don't need such capabilities Let the local cloud providers focus on the subset of "solved" infrastructure (VM, Storage, DNS, CDN, DB Hosting). And of course there is nothing stopping a BD startup from storing their data on a local cloud provider and then transferring subsets of this data to a big cloud provider for AI/ML processing on an as-needed basis.

This is better for the country in the long run. It lets the people develop cloud services from the ground up and truly understand how it all works from the fundamental building blocks - rather than skip this process and go to a big provider who has solved all the challenges. You should not be proud that India has opted to skip this in favor of AWS. It robs your country-men of a truly valuable learning experience that would only enhance their overall technical competency.

আপনার বাড়ি কোথায় ভাই ? কোন জেলায় ?

 

[mb444]

You don't understand. At the end of the day it doesn't matter if big cloud providers provide objectively better service. There is no big culture of cloud in BD yet. Use of certain software tools/infrastructure can be cultural - just look at Japan or China (AliCloud). So rather than let the growing BD IT professionals fall into using the international cloud providers - we can instead use this moment to re-direct them to use local Cloud services. Of course the onus is on the local cloud services to provide outreach and training on their offerings - and there should be a strong culture of feedback so they can tailor their services and support specifically to the needs and workflows of Bangladeshi IT professionals (mostly web-devs).

The big cloud providers can handle "advanced services" - aka AI/ML workloads. 90% of apps don't need such capabilities Let the local cloud providers focus on the subset of "solved" infrastructure (VM, Storage, DNS, CDN, DB Hosting). And of course there is nothing stopping a BD startup from storing their data on a local cloud provider and then transferring subsets of this data to a big cloud provider for AI/ML processing on an as-needed basis.

This is better for the country in the long run. It lets the people develop cloud services from the ground up and truly understand how it all works from the fundamental building blocks - rather than skip this process and go to a big provider who has solved all the challenges. You should not be proud that India has opted to skip this in favor of AWS. It robs your country-men of a truly valuable learning experience that would only enhance their overall technical competency.

Your points are well made regarding internet based apps or microservices leveraging cloud solutions.

However true ML/AI/deep learning projects that truly leverages elastic compute capacities of cloud computing deals in petabytes of data or high volume streamed data or a combination of the two. The sources of such data is usually financial transactions, factory production process type data, sales and marketing data, scientific data such as genome or facial or some kind of pictorial/video/audio data.

These type of projects rarely in my experience leverages the services provided by big players. The latest packaged product such as microsoft azure synapse or AWS Glue which supposedly can ingest data, undertake etl or process data data under AI/ML scenario and egress it. But they are just awful and do not work. Hell neither platform has yet released an end-to-end job monitoring service.

Every company I know builds their own bespoke AI/ML models or leverages specialised vendors such as informatica or sector specific niche products.

At the end of the day data lakehouse/mesh are still very much conceptual methodologies that are evolving.

BD companies based on market capitalisation presumably are firmly at the stage where they are using on-prem ERP products and at the clasp where volumes of data is getting to a point where data lakes becomes the only option to truely leverage their data.

Here if digital BD is seriously the goal then we need to develop indeginious solutions. The beauty of decentralised computing is that the framework is open source. Hardware setup is also fairly simple.

It is an absolute fallacy that we need to rely on established players. BD is making some right moves and i am hopeful of the future.

Попередження про переадресування

www.google.com

Попередження про переадресування

www.google.com

Simple google search reveals many established AI/ML consultancies already exists in country.

Top 20+ Artificial Intelligence Companies in Bangladesh (2025) - TechBehemoths

Discover the best Artificial Intelligence companies in Bangladesh. 34 companies are available in this region. Hire the top Artificial Intelligence company in Bangladesh for your project!

techbehemoths.com

 

[sonargon]

Your points are well made regarding internet based apps or microservices leveraging cloud solutions.

However true ML/AI/deep learning projects that truly leverages elastic compute capacities of cloud computing deals in petabytes of data or high volume streamed data or a combination of the two. The sources of such data is usually financial transactions, factory production process type data, sales and marketing data, scientific data such as genome or facial or some kind of pictorial/video/audio data.

These type of projects rarely in my experience leverages the services provided by big players. The latest packaged product such as microsoft azure synapse or AWS Glue which supposedly can ingest data, undertake etl or process data data under AI/ML scenario and egress it. But they are just awful and do not work. Hell neither platform has yet released an end-to-end job monitoring service.

Every company I know builds their own bespoke AI/ML models or leverages specialised vendors such as informatica or sector specific niche products.

At the end of the day data lakehouse/mesh are still very much conceptual methodologies that are evolving.

BD companies based on market capitalisation presumably are firmly at the stage where they are using on-prem ERP products and at the clasp where volumes of data is getting to a point where data lakes becomes the only option to truely leverage their data.

Here if digital BD is seriously the goal then we need to develop indeginious solutions. The beauty of decentralised computing is that the framework is open source. Hardware setup is also fairly simple.

It is an absolute fallacy that we need to rely on established players. BD is making some right moves and i am hopeful of the future.

Попередження про переадресування

www.google.com

Попередження про переадресування

www.google.com

Simple google search reveals many established AI/ML consultancies already exists in country.

Top 20+ Artificial Intelligence Companies in Bangladesh (2025) - TechBehemoths

Discover the best Artificial Intelligence companies in Bangladesh. 34 companies are available in this region. Hire the top Artificial Intelligence company in Bangladesh for your project!

techbehemoths.com

Yes agree with you on AI/ML for cloud. The true reality is that I do not know anyone who uses the big provider AI/ML services - they only rent the GPU compute and then build their own AI/ML pipelines on top of those. So there is definitely room to compete in this area.
I have built data-lake-house in AWS for financial transaction data (storage and analytics) - all "serverless" - and yes its not easy or simple at all. The pre-packaged offerings suck and you have to design your own event based architectures to ingest data into s3. AWS Glue sucks so much balls. And all these "data-lake-houses" are just means to store data in some way so that they can be queried by Presto, Hive, Spark, etc (all open-source!). So definitely these offerings can be provided by a local provider if they put in the effort. In many ways the bar is quite low!

Ideally BD is able to pioneer indigenous solutions! Its not out of reach at all. But it will require concerted efforts. Imagine a country where the local cloud provider(s) have strong relationships with local software firms. None of the BS with AWS or Azure who juggle hundreds of multi-billion dollar MNC's and don't give a F about you. But a local provider who would have every reason to provide good support to keep your business. And they would have every reason to consistently innovate and offer new services and workflows demanded by local customers. There would be lots of collaboration between both parties and immense knowledge transfer.

And it would create a great tech talent eco-system in BD. On the cloud provider side - Engineers are needed for data-center design, IT professionals are needed for server administration and hardware virtualization (splitting 1 server into multi-tenant, kubernetes, etc). Software developers are needed for API and Web interface development/implementation. Furthermore systems analysts, hardware technicians, customer service representatives, account executives, and finance teams are needed for operations. And there would be immense talent flow between local cloud providers and local software firms.

আপনার বাড়ি কোথায় ভাই ? কোন জেলায় ?

Why do you ask? I do not know you so I will not say.

 

[Pranaayam]

No. Every SLA that likes of Azure have revolves around data redundancy or how many copies of the data one holds. Storage is first priority.

Then you move onto processing data. Here you have to build your own data processes using databricks or whatever you want. There are slas around uptimes of cluster and how elastic the resource provisioning will be.


You are mistaking cloud with an ERP system the like of sap or oracle that comes with best of breed data models and processes. No such things exist in cloud.

There is nothing BD need do that many countries have not already done. Big providers are not necessary when it comes down to it. Its fine to have your view but is not something that many policy makers would share. You serve the interest of your country not foreign IT companies.

I don't have a skin in the game. I have my opinion coming from 15 years as an IT specialist. I have seen my share of companies evaluating clouds and giving up on markets since they don't fit in the larger scheme of things(read revenue). I am not optimist about Bangladesh forcing anyone on data localisation. Let's agree to disagree. Like I said, I have no skin the game, it is for your policy makers to take the call.

Yes agree with you on AI/ML for cloud. The true reality is that I do not know anyone who uses the big provider AI/ML services - they only rent the GPU compute and then build their own AI/ML pipelines on top of those. So there is definitely room to compete in this area.
I have built data-lake-house in AWS for financial transaction data (storage and analytics) - all "serverless" - and yes its not easy or simple at all. The pre-packaged offerings suck and you have to design your own event based architectures to ingest data into s3. AWS Glue sucks so much balls. And all these "data-lake-houses" are just means to store data in some way so that they can be queried by Presto, Hive, Spark, etc (all open-source!). So definitely these offerings can be provided by a local provider if they put in the effort. In many ways the bar is quite low!

Ideally BD is able to pioneer indigenous solutions! Its not out of reach at all. But it will require concerted efforts. Imagine a country where the local cloud provider(s) have strong relationships with local software firms. None of the BS with AWS or Azure who juggle hundreds of multi-billion dollar MNC's and don't give a F about you. But a local provider who would have every reason to provide good support to keep your business. And they would have every reason to consistently innovate and offer new services and workflows demanded by local customers. There would be lots of collaboration between both parties and immense knowledge transfer.

And it would create a great tech talent eco-system in BD. On the cloud provider side - Engineers are needed for data-center design, IT professionals are needed for server administration and hardware virtualization (splitting 1 server into multi-tenant, kubernetes, etc). Software developers are needed for API and Web interface development/implementation. Furthermore systems analysts, hardware technicians, customer service representatives, account executives, and finance teams are needed for operations. And there would be immense talent flow between local cloud providers and local software firms.



Why do you ask? I do not know you so I will not say.

Do you think your local apps will warrant a separate cloud services provider? Does it make economic sense. If so, why not put money where you mouth is and start this yourself.

You seem to think a country like Bangladesh can manage to develop a complete cloud solution on its own. I hope you'll invest your money into such a non starter. It's easy to say everything is easy. AMAZON invested billions and a decade of time to get to where it is now. You are overestimating your country's skill to accomplish something like this.

Thankfully India is smart enough to not reinvent the wheel. Luckily, INDIA is a large enough and economically important enough for all cloud providers to set up data centers in the country. As such, INDIA is more integrated than isolated. INDIA has such huge IT imprint that most if not all software in the world has some indian involvement. India is able to do so because it understands its priorities and don't go about wasting time and effort on useless endeavors.

 

[mb444]

I don't have a skin in the game. I have my opinion coming from 15 years as an IT specialist. I have seen my share of companies evaluating clouds and giving up on markets since they don't fit in the larger scheme of things(read revenue). I am not optimist about Bangladesh forcing anyone on data localisation. Let's agree to disagree. Like I said, I have no skin the game, it is for your policy makers to take the call.

Do you think your local apps will warrant a separate cloud services provider? Does it make economic sense. If so, why not put money where you mouth is and start this yourself.

You seem to think a country like Bangladesh can manage to develop a complete cloud solution on its own. I hope you'll invest your money into such a non starter. It's easy to say everything is easy. AMAZON invested billions and a decade of time to get to where it is now. You are overestimating your country's skill to accomplish something like this.

Thankfully India is smart enough to not reinvent the wheel. Luckily, INDIA is a large enough and economically important enough for all cloud providers to set up data centers in the country. As such, INDIA is more integrated than isolated. INDIA has such huge IT imprint that most if not all software in the world has some indian involvement. India is able to do so because it understands its priorities and don't go about wasting time and effort on useless endeavors.

We are fundamentally speaking about two different things.

If we are talking about ERP solution on-prem or cloud based then absolutely i am in agreement with you. BD companies should use best of breed industry solutions such as SAP, Oracle or Salesforce. There is no value in seeking to reinvent the wheel.

However cloud computing that brings together huge disparate datasets the only solution is datalake. Here no particular advantage is held by big players at all. All you need are data centres and compute powers. If business is there DC will be set up by big players or like in china they will cooperate with local players to provide their service ( in china MS or Amazon can not trade under own name. They do not have DCs. They partner with local DCs to provide service)

Then you have AI/ML projects. Again these tech is evolving. At the moment standardisation has not occurred and big players are being outplayed by niche players.

Some on us here are commenting on these scenarios that requires materially different policy response. Issues like geo fencing data does not occur for ERPs because globally very few companies use entirely cloud based solutions.

 

[sonargon]

Do you think your local apps will warrant a separate cloud services provider? Does it make economic sense. If so, why not put money where you mouth is and start this yourself.

You realize that even in the west many software firms forgo Cloud Service providers to architect their own bespoke solutions on co-located clusters? It always ends up being cheaper and more effective to do this. Cloud is still mostly preferred by large MNC's who can throw money at it and care more about SLA and having someone to blame when there are outages. "Nobody ever got fired for buying IBM". I already told you there is nothing extraordinary cloud providers are offering - you will see many debates among tech professionals on whether moving to the cloud truly resulted in a net benefit. It is a move mostly preferred by finance people who see it as a way to cut operational hiring costs because it removes the need to hire Sys Admins and IT professionals and instead shifts their work onto software developers.

I will not respond to you further because you are dumb and fail to understand my point.

Thankfully India is smart enough to not reinvent the wheel. Luckily, INDIA is a large enough and economically important enough for all cloud providers to set up data centers in the country. As such, INDIA is more integrated than isolated. INDIA has such huge IT imprint that most if not all software in the world has some indian involvement. India is able to do so because it understands its priorities and don't go about wasting time and effort on useless endeavors.

I laugh at this. I see very few contributions to open source by Indian developers. Majority of it is from Europeans and Americans. I see more contributions from Iranian's then indians. Those who contribute to open source leave an imprint on software not masses of Java enterprise programmers.

The indian "imprint" on IT is garbage work done by WITCH companies and youtube tutorial videos for Java lol.