As a Dhaka-based IT services company, we help international companies with six services covering the whole life of a system: building it, connecting it to what you already run, modernizing it when it ages, testing it, and keeping it alive afterwards. Code lives in your repositories, and every project ends with a handover your own engineers can use.
Book a 30-minute scoping call See the six services
Built in your own repositories, cloud and ticketing · Confidentiality, data processing and intellectual property terms signed before any code · Overlap hours agreed at kickoff
Our IT services cover six engagements across the life of a system, from first build to years of upkeep. Each has its own scope and price; most clients buy two or three together, such as a build with testing, or an API integration with maintenance. Staffed level 1 and level 2 user support runs separately as the outsourced help desk.
Custom web and backend systems built to a written spec, with tests, code review, continuous integration and documentation your engineers can take over.
Web and mobile applications from wireframe to release, built in your repositories with staged environments and a written release process.
Your systems connected to payment, sales, finance, logistics or government APIs, with retries, error handling and monitoring on every call.
Legacy systems assessed, then moved off dead frameworks in stages, with the business running throughout rather than in a risky rewrite.
Bug fixes, dependency and security updates and monitoring under a monthly agreement with an agreed response time per severity level.
Test plans, automated regression suites and release checks, so a deployment stops breaking things your team only finds in production.
Our IT services work across the life of one system in five stages: build it, connect it, test every release, keep it running and modernize it when it ages. Each stage has its own service and its own price. All of them work in your repository from the first commit.
The IT service you need first depends on where the software is now. Nothing built yet means software development. Built but disconnected means an API integration. Old and fragile means application modernization. Live and breaking means testing and maintenance, and those two protect revenue you already earn.
| Your situation | What you need | Where we start |
|---|---|---|
| You run the process on spreadsheets and email | Software Development | Scope and estimate |
| Customers or staff need a web or mobile app | Application Development | Wireframe review |
| Two systems hold the same data and nobody syncs them | 3rd Party API Integration | Integration audit |
| The stack is old and nobody wants to touch it | Application Modernization | Codebase assessment |
| Releases keep breaking things you find later | Software Testing | Test plan |
| Nobody owns updates, patches or the on-call phone | Maintenance and Support | Support agreement |
| Everything already runs, is tested and is documented | No action needed | Nothing yet; talk to us when something breaks |
Software is built and handed over under three rules Eicra Soft Limited calls inheritor-first engineering, applied in this order: build for the engineer who inherits it, integrate through APIs rather than people, and modernize in stages. The rules follow the NIST SSDF practice groups, and we say so when the cheaper option is the right one.
Software is cheap to write once and expensive to change forever. The cost that matters is not the build; it is the next five years of someone else reading it.
So every project ships with tests, a setup guide that actually starts the project, environment documentation and a second named reviewer on every change.
If a person exports a file on Monday so another system has Tuesday’s data, that is not an integration. It is a job that fails quietly whenever that person is on leave.
So we connect systems directly, with retries, an error queue, an alert to a named owner and a log of every call that failed and why.
Full rewrites fail for a predictable reason: the old system keeps changing while the new one is being built, and the business cannot wait for the finish line.
So we move one module at a time behind a stable interface, keep both running, with a rollback ready, until the new path is proven, and stop whenever you want to stop. The pattern is the strangler fig application described by Martin Fowler.
An offshore IT services company earns its fee after launch, when the code has to be read, changed and secured by someone else. We build to the NIST Secure Software Development Framework, review every change with a second named engineer and leave tests your own team can run.
| Item | Without | With EICRA |
|---|---|---|
| Scope | “Let us start and see how it goes” | A written spec with named screens, endpoints and acceptance tests |
| Price | Hourly, discovered as you go | Fixed for projects; per engineer per month for continuous work |
| Code review | Whoever wrote it merged it | Reviewed by a second named engineer, on every change |
| Testing | Manual clicking before release | Automated regression tests that run on every deploy |
| AI-written code | Merged because it runs | Reviewed against the Open Worldwide Application Security Project (OWASP) Top 10 before it ships |
| Handover | A zip file and a phone call | Your repository, your cloud, environment docs and a runbook |
| After launch | Support billed per request | A monthly agreement with a response time per severity level |
| Contracts | A freelancer invoice | Company contract with a non-disclosure agreement (NDA), a data processing agreement (DPA) under General Data Protection Regulation (GDPR) Article 28(3), and intellectual property (IP) assignment |
These answers cover what international buyers ask Eicra Soft Limited before a scoping call: how IT services are priced, who owns the code, how a project starts, what happens after the call, how a takeover from another vendor works, where code and data stay, and what these six services deliberately leave out. Each answer stands on its own.
Reviewed by Asma Sultana Minu, Local Corporate Counsel.
More answersIT projects follow a fixed ladder: a $1,500 diagnostic, a $4,000 pilot and production from $6,000. Continuous work is $4,200 per engineer per month ($4,800 senior) with a three-month minimum. Most engagements open with a two-week software diagnostic whose fee is credited to the pilot that follows.
An IT services company builds, connects, tests and maintains the software a business runs on, and often manages the servers and networks underneath. Unlike a product vendor, it works on your own systems under contract, with named engineers, a written scope and a handover your team can use.
Buy fixed-price projects when the work has a clear end, and hire a dedicated development team when the backlog keeps growing. Projects are quoted once; a dedicated team is billed per engineer per month with a three-month minimum, at a rate agreed in the written scope.
You do, and it is assigned in writing rather than assumed: source code, build configuration, designs and documentation. As a Bangladesh-based team we work in your repositories from the first commit, with named, least-privilege access removed at handover. Open-source components keep their own licences, listed before delivery.
Usually within one to two weeks of a signed scope, beginning with the two-week diagnostic. Repository access, environment credentials and approvals cause most delays, so the contract lists every login, environment and named approver we need before any work starts.
After the scoping call, the work moves in four written steps before any code is committed:
Yes, taking over software another vendor built is a normal starting point, and it begins with a codebase assessment rather than a rewrite. The assessment inventories the repositories, environments, dependencies and undocumented behaviour, then produces a written list of what must be fixed, tested or documented before new work is safe. Where the previous vendor still holds code or credentials, we list exactly what to request from them before the handover date.
Your code and data stay in your own repositories, cloud accounts and ticketing system for the whole engagement; nothing is copied to infrastructure we control. Access is named and least-privilege, and it is removed at handover. Personal data is covered by a data processing agreement under GDPR Article 28(3), and the AI coding tools permitted to touch your code are agreed in writing at kickoff, alongside the overlap hours.
These six IT services do not cover server, network or end-user helpdesk operations, and we do not start work without a written scope. Day-to-day user support for your ticketing system is a separate managed service, the outsourced help desk, and staffing for an open-ended backlog is a dedicated development team rather than a fixed-price project. If a cheaper option than ours is the right one, we say so during the scoping call.
Tell us what breaks most often. We will tell you whether it needs a fix, a test or a rebuild.
Request A QuoteBook a 30-minute scoping callYour request is subject to our Terms and Privacy Policy.
