Senior software engineering for the systems your business runs on.
We design, build and scale payments, integrations, data platforms, web and mobile apps and applied AI for enterprises and scaling companies. Led by Saidul Islam Bhuiyan, with a team of full-stack and UI/UX developers and 13+ years shipping for Time Doctor, Wego, Zenius and Exporo.
09:40:55rota.draft.createdwaiting for coordinatorreview
Illustrative events, modeled on systems we have shipped.
Senior-led teamFull-stack and UI/UX developers
$18Ma year in savings the Time Doctor billing rebuild contributed to
8M+registered users on Zenius, scaled as Staff Engineer
2M+users on the Wego hotel metasearch, fed by 40+ suppliers
280K+active users on the Time Doctor billing service
Where are you today?
Every product starts somewhere. Pick where you are.
Enterprise teams, scaling products and new ventures need different things from a technical partner. Here is where we would start with you, and what you would walk away with.
You need a product people will pay for, without spending the budget on the wrong features.
Where we start
A short discovery to pin down your first users, the one workflow that must work on day one, and how the product makes money. Then a scoped plan and a first working release.
What you walk away with
A scoped MVP with clear milestones and budget
Web and mobile apps your first users can sign up to and pay for
An architecture chosen so version one can grow without a rewrite
Proof
Accordium. The first version of a sales automation SaaS, part of the product readiness behind its Series B.
Agents, MCP servers and RAG that people use every day, with a safety check or a human approval before anything reaches a user.
Proven at
Everhome
Ovue ANCORE
ThemeXpert
AI agents and copilots
Agents that match people, draft rotas and write daily briefs, with people approving the result.
GeminiAI agents
MCP servers
Your systems exposed as typed tools, so teams can ask operational questions in plain language.
MCPNode.js
RAG and document AI
Answers grounded in your own content, and readers that turn documents into checked fields.
RAGVector searchEmbeddings
Guardrails and human approval
Server side safety checks and approval steps before any output reaches a user.
Safety checksAudit trail
Architecture
AI agent and MCP server architecture, with a safety check on every answer and human approval for changes. Hover or tap a step to see it in the diagram.
1
AI agentPlans the steps and calls tools to answer, on models like Gemini.
2
MCP serverExposes your systems as typed tools, with authentication and an audit log.
3
Your data and documentsTools query databases and APIs, and RAG answers from your own documents.
4
Safety checkEvery answer is checked on the server before a user sees it.
5
Human approvalAnything that changes data waits for a person to approve it.
Built for Everhome coordinators and Ovue ANCORE crews.
Production systems for companies in the USA, Europe and Asia.
Two deep dives, then a wider selection. Each shows the role held and what it changed for the business. Numbers come from the clients' own published figures or from the project itself.
Moving subscription billing for 280K+ users from Stripe to Paddle while the business kept running.
Context
Billing lived inside a legacy service tightly coupled to Stripe. Subscriptions and churn were hard to manage, and partner integrations depended on the same code.
Constraints
Live revenue for 280K+ active users
Partner and third party integrations had to keep working
Every subscription change had to be traceable
Key engineering decisions
Billing split into NestJS microservices that communicate through GCP Pub/Sub events, so subscription, churn and partner flows can change independently
Every Paddle webhook handled idempotently, so a retried delivery never applies twice
An auditable event trail for each subscription and churn change
Legacy partnership services and their data moved onto the new platform with minimal disruption
Infrastructure in Terraform, with 24/7 monitoring and alerts in Datadog
Results
280K+active users on the new billing
9 monthsto go live
$18Ma year in savings it contributed to
Simplified architecture. Service names are descriptive.
Lead Backend Engineer, Hotel Integrations at GoQuo and Wego
One real-time hotel search across 40+ suppliers for 2M+ users.
Context
Wego added hotels next to its flight search. The new product had to compare rates from 40+ suppliers, including Booking.com, Agoda, Traveloka and Hotelbeds, and always show the cheapest available price.
Constraints
Every supplier with its own API and data format
Search by map location and by hotel or city name
Prices that change constantly
Key engineering decisions
Scheduled ETL brings every supplier feed into one rate model
Elasticsearch handles geo and full text search
Price changes reach users live through Redis and WebSockets
Services in Go and Node.js talk over gRPC and NATS, running on Kubernetes in GCP
Supplier integrations kept running through the product’s move into Wego
Results
2M+users on hotel search
40+supplier integrations
Liveprices pushed as rates change
Simplified architecture. Service names are descriptive.
AI care operations platform with a human sign off on every decision
One backend runs the office, carer and family apps. AI matches carers and drafts the rota for coordinators to approve, onboarding runs on Vouchsafe, and visits flow into payroll and Stripe invoicing.
Business resultAdmin work drafted by AI and approved by people, with agencies kept ready for CQC inspections.
3 mincarer onboarding checks, down from hours of paperwork
Node.jsFastifyAzureSQL ServerStripeMCP
EdTech · Indonesia
Zenius
Staff Engineer
Learning platform and infrastructure at national scale
Co-designed a microservices LMS connected through Kafka, SNS and WebSockets, and led the schema, indexing, sharding and caching work that kept performance steady and brought cost down.
Business resultSteady performance and lower infrastructure cost as the platform grew, plus the technical plan for Zencore, a new gamified product.
8M+registered users
Node.jsKafkaPostgreSQLMongoDBGraphQL
Real estate fintech · Germany
Exporo
Senior Software Engineer
Fiat and crypto payments as auditable state machines
Investors fund property projects in fiat or crypto through MangoPay and Upvest. Each payment moves through an XState state machine on AWS serverless, so every transition is recorded.
Business resultInvestors could fund projects with crypto as well as fiat, which opened a new source of revenue.
2 railsfiat and crypto in one payment service
TypeScriptXStateLambdaStep FunctionsDynamoDB
Maritime wellbeing · Wearables
Ovue ANCORE
Senior Engineer in a small team
Smartwatch sync and AI health insights for crews at sea
A Kotlin Bluetooth module keeps watch data syncing with no internet, an offline queue uploads small daily snapshots over slow ship links, and Gemini insights pass a safety check before a seafarer sees them.
Business resultCrew health data keeps flowing with no internet at sea, and both crew modules shipped for the ANCORE pilot.
Offlineby design, delivered for the ANCORE pilot in Sept 2026
React NativeKotlinBLEGeminiCloud Run
Public transport · Bangladesh
BRTC
Full-stack Lead
Multi-tenant ticketing for the national bus operator
Manual cash tickets replaced by Android POS terminals across the fleet, a NestJS backend with batch ticket sync, and a React admin where each depot sees only its own data.
Business resultFares recorded digitally instead of in cash, with each depot seeing its own numbers.
POSticketing on Sunmi terminals across the fleet
NestJSReactAndroidMongoDBSocket.IO
Biometric security · South Korea
Suprema BioStar 2
Performance and test automation
Fast queries over a terabyte of access logs
Logs were partitioned by day and month and indexed in MariaDB and SQL Server, the Node.js backend moved to async processing to cut memory leaks, and a Karate suite kept live customer sites safe.
Business resultFast queries for live customer sites, and releases that stay backward compatible.
1TB+time-series access logs tuned and migrated
Node.jsMariaDBSQL ServerKarate
Also delivered
Telekom MalaysiaCorporate website rebuild with an interactive fibre coverage map
BigWIFI, CanadaCustomer platform with speed tests, live chat and a tech support marketplace that opened a new revenue stream
ThemeXpertAI plugin for WhatsApp, WooCommerce and Facebook with RAG over site and PDF content
MovingwallsData driven booking platform for digital billboards across Malaysia
AccordiumFirst version of a sales automation SaaS, part of the readiness behind its Series B
CapitalDKVoice driven prototype for oil and gas field operations
Client feedback
What clients and colleagues say.
In their own words, from clients in the USA, the UK and Malaysia, and an engineer who built alongside Saidul at Zenius.
“
Saidul Islam Bhuiyan Technology was one of the best tech leads and engineers I’ve ever dealt with. They were ethical, hardworking, and really skilled. Their communication was also great.
At Zenius, I and Saidul worked together on modules related to authentication and payments and some more. Saidul brings great design and ideas to the table and I have personally learned a lot from his abilities. We have collaboratively worked and learned different tech stacks from each other. He is one of the most sorted engineers I have worked with. Would love to work with him again in the future.
AS
Akhil S.Engineer who worked with Saidul at Zenius · India
“
Saidul knows how to work and communicate efficiently. He takes full responsibility of what he does and always exceeds my expectation.
TP
Terence P.Client · USA
“
Had a great experience with Saidul, during which he far exceeded expectations on this small project. Thank you, Saidul!
N
NickClient · United Kingdom
“
5 Star overall, deliver quality work as promised.
BL
Bryan L.Client · Malaysia
Partnership
A partner in your product, not another agency.
Saidul leads every engagement and stays your technical point of contact, backed by a team of full-stack and UI/UX developers. Success is measured the way you measure it, in users, revenue, uptime and running cost.
TopicTypical agencyWorking with us
Who you talk toAn account manager passing messages alongThe senior engineer designing and building your product
Who builds itWhoever is free, often junior developersSaidul leads, with a team of full-stack and UI/UX developers
How scope is handledThe ticket gets built as writtenWe push back when a simpler path reaches users and revenue sooner
Who owns the codeRepos and cloud sit in the agency accountsYour repos, your cloud accounts and your IP from day one
How you payLarge deposits or retainers paid in advanceMilestone billing under your MSA and SOW, or escrow if you prefer
After launchA final invoice and a goodbyeA 10–60 day warranty, a 24 hour response, then ongoing support
How success is measuredHours billed and tickets closedYour users, revenue, uptime and running cost
Clients come backExporo and other clients hired again after the first contract.
Trusted with other teams' codeLong-running roles as senior code auditor and technical PM for clients on Upwork.
Expert-Vetted on UpworkIndependently vetted into the top 1% of talent, with verified client feedback on the profile ↗
Product thinking
Six questions we answer with you before any code is written.
They decide whether a product earns money or only ships.
?
Who pays, and what problem makes them pay?Features are chosen around the buyer and the problem they pay to solve.
?
What is the one workflow that must work on day one?That becomes the first release. Everything else can wait.
?
What breaks first at ten times the users or data?Planning for it now costs far less than a rewrite later.
?
What does it cost to run per customer?Cloud, AI and third party fees decide your margin.
?
Where can AI save time without putting users at risk?AI drafts, people approve, and answers are checked before users see them.
?
How will we know it is working?We agree the numbers that matter, such as sign ups, conversion or hours saved, before launch.
Prototype to production
Built it with AI? We make it ready for real users.
Lovable, Figma and AI coding tools prove an idea fast. When real users, data and features arrive, we take over where AI development stops scaling and turn the prototype into production web and mobile software.
Left, the shape most AI built prototypes share. Right, what they become. Logic moves behind an API, slow work goes to a queue, AI output is checked, and every change ships through tests and CI.
Signs your prototype has hit its limit
Each new feature breaks something that used to work
Prompting fixes one bug and creates another
Pages slow down and requests time out as data grows
Auth, payments and data access you are not sure are secure
No tests, no staging and no safe way to roll back
Hosting and AI bills rising faster than usage
The same takeover work, on larger systems
Suprema BioStar 2Took over a live platform straining under 1TB+ of logs and made it fast, async and covered by tests.
Time DoctorReplaced a tightly coupled payment service with an event-driven one for 280k+ active users.
ZeniusRedesigned schema, indexing, sharding and caching as usage grew to 8M+ registered users.
A clear path from first call to a system your team owns.
At every phase you can see what we do, what you approve, and when money moves.
Each column is a phase. You approve the plan and accept each milestone, and each acceptance releases that milestone’s payment. After launch, the warranty runs at no cost.
Engineering and security practices
Practices a CTO can check against real systems.
Each one comes from a production system described on this page.
NDAs, IP assignment and your security questionnaire are handled before kickoff.
Idempotent, auditable money flowsDeduplicated webhooks at Time Doctor and payment state machines at Exporo, so every transaction can be traced.
Migrations with minimal disruptionLegacy partner services and their data moved onto new platforms while customers kept working.
Backward-compatible releasesRegression suites, like the Karate suite at Suprema, protect live customer sites from breaking changes.
Observability and infrastructure as codeTerraform, dashboards and 24/7 alerting in Datadog go live with the system.
Access control and data isolationRole based access, tenant and depot level isolation, and permanent audit trails.
AI behind guardrailsA human approval or a server side safety check before any model output reaches a user.
Ways to work together
Find the engagement that fits your team.
Project delivery, senior engineers inside your team, or an architecture review. Answer three questions to see where we would start.
Answer three questions
Do you have engineers in-house?
Is there a live system or prototype already?
Is the scope already clear?
Where we would start
Architecture review→Embedded engineers
We review the system and agree a plan, then senior engineers join your team to carry it out.
First step.A focused review with a written report and a prioritized plan.
Every engagement starts with a discovery and architecture sprint.
The 1 to 2 week sprint costs $1,000 to $2,000 USD and ends with an architecture plan, a scoped roadmap and a delivery budget you sign off. A typical delivery runs up to $30,000 to $40,000 USD depending on requirements and complexity, and larger programs are scoped and quoted separately.
Billing follows the milestones in your SOW. A small upfront payment starts the project, and each later payment is due only when its milestone is reached.
Invoices that fit your MSA, SOW and procurement process
Escrow available for extra protection
Every payment tied to a milestone you can see and test
Project startsA small upfront payment
Milestone 1 reachedInvoiced, or released from escrow
Milestone 2 reachedInvoiced, or released from escrow
Final deliveryLast payment, then the warranty period begins
Example plan. The number of milestones depends on the project.
Included with every project
A 10 to 60 day warranty, with a response within 24 hours.
If a bug or issue shows up after delivery, we fix it at no extra cost for a minimum of 10 to 60 days, with the exact window written into your plan before work starts. Support requests get a response within 24 hours.
10–60day warranty
24hresponse time
Stack and industries
Tools chosen for the problem, used in production.
Node.js and TypeScript at the core, Go where throughput and latency matter, on AWS, GCP or Azure.
SaaSPayments and FintechHealthcare and Home CareMaritimeTravelEdTechMedia and AdvertisingOil and GasReal Estate InvestmentTelecomPublic TransportBiometric SecurityE-commerce
Who you work with
SB
Saidul Islam BhuiyanFounder and principal engineer
Senior backend engineer with 13+ years of building and scaling production platforms. Saidul has worked as Staff Engineer at Zenius, Lead Backend Engineer at GoQuo and Wego, and Integrations Engineer at Time Doctor, and now leads a team of full-stack and UI/UX developers working with clients across Europe, North America and Asia.
The work usually covers the whole path, from architecture and data model to infrastructure, monitoring and handover, alongside mentoring the engineers who will carry it forward.
No. Saidul leads every engagement and stays your technical point of contact, backed by a team of full-stack and UI/UX developers, so work keeps moving when one person is away.
Who owns the code and the IP?
You do. The code lives in your repositories and runs in your cloud accounts from day one, and the intellectual property is yours.
Will you sign an NDA and complete our security review?
Yes. We sign NDAs before you share anything sensitive, and can complete your security questionnaire before kickoff.
How do contracts and payments work?
Work runs under an MSA, with a statement of work for each phase. A small upfront payment starts the project, and every other payment is due only when its milestone is reached. Escrow is available if you prefer.
How will I know how the project is going?
You see working software at every milestone, with a short written update on progress, risks and next steps.
What support is included after launch?
A minimum 10 to 60 day warranty after delivery, so bugs and issues are fixed at no cost, and support requests get a response within 24 hours. After that we can stay on for support and new features, or hand over with documentation and a walkthrough so your team can run it.
Can you work with our in-house team?
Yes. We can lead the build, join your sprints as senior engineers, or review your architecture and mentor your engineers.
Which time zones do you cover?
We are based in Tangier, Morocco (GMT+0), with full overlap with UK and European working hours and the morning of the US East Coast.
How much does an engagement cost?
Engagements start with a 1 to 2 week discovery and architecture sprint that costs $1,000 to $2,000 USD. A typical delivery runs up to $30,000 to $40,000 USD depending on requirements and complexity, and larger programs are scoped and quoted separately.
We built an app with Lovable or another AI tool. Can you take it over?
Yes. We audit it first and tell you plainly what to keep, refactor or rebuild, then take it to production on a backend built to scale.
We only have an idea. Can you help?
Yes. A discovery sprint defines your first users, the core workflow and how the product makes money, then we build a first release you can put in front of customers.
Do you build AI agents and MCP servers?
Yes. We build AI agents that call your systems through MCP servers with typed tools, authentication and an audit log. At Everhome an MCP server lets coordinators ask operational questions in plain language, and every answer we ship passes a safety check first.
Can you build Bluetooth Low Energy apps that work offline?
Yes. For Ovue ANCORE we built a Kotlin Bluetooth LE module with a serialized GATT queue and backoff reconnects, plus an offline queue that uploads small batches whenever a ship gets a connection.
Which cloud platforms do you work with?
AWS, Google Cloud, Azure and DigitalOcean. Infrastructure is written in Terraform, services ship as Docker images, and monitoring runs in Datadog and Sentry.
Do you work with startups or only enterprises?
Both. Enterprises and scaling companies are most of our work, and founders with an idea or an AI built prototype start with the same discovery sprint.
Start a project
Tell us about the system you need to build or scale.
Book a call, or send a short brief. You will get a straight answer on scope, approach, timeline and whether we are the right fit.
NestJS
Fastify
Go
PostgreSQL
Kafka
Time Doctor
Wego
Zenius
Exporo
Suprema BioStar 2
Telekom Malaysia
Everhome
Ovue
Senior-led teamFull-stack and UI/UX developers
Stripe
PaddleMangoPay
XState
ElasticsearchWebhooks
Redis
MariaDB
SQL Server
AWS
GCP
DigitalOcean
Terraform
Docker
Datadog
Sentry
Kubernetes
Movingwalls
BRTC
FigmaUser flowsRoadmap
React
Angular
TypeScriptWebSockets
Kotlin
Android
Bluetooth LE
MQTT
MongoDB
ThemeXpert
GeminiAI agents
MCP
Node.js
NATS
Azure
GraphQL
DynamoDB
Cloud Run
Socket.IO
BigWIFI, CanadaCustomer platform with speed tests, live chat and a tech support marketplace that opened a new revenue stream
AccordiumFirst version of a sales automation SaaS, part of the readiness behind its Series B
CapitalDKVoice driven prototype for oil and gas field operations
Trusted with other teams' codeLong-running roles as senior code auditor and technical PM for clients on Upwork.
Expert-Vetted on UpworkIndependently vetted into the top 1% of talent, with verified client feedback on the profile ↗
JavaScript
Python
Java
PHPSQL
Express
Spring BootREST
MySQL
Prisma
RabbitMQ
Payments and Fintech
Healthcare and Home Care
Maritime
Travel
EdTech
Media and Advertising
Real Estate Investment
Telecom
Biometric Security
E-commerce
LinkedIn ↗
Freelance senior engineer and leadEverhome, Ovue, BRTC, Suprema, ThemeXpert