I'm a Freelance Senior Software Engineer based in Dubai, UAE.
Let's work togethersameerizwan3@gmail.comDubai, UAE
From monolith to microservice, from spreadsheet to pipeline. I design backends for banks that cannot go down and AI systems that do the work nobody has time for — then prove it with numbers, not adjectives.
500K+
Live users on the digital onboarding platform I engineered for Banque Misr — across roughly twelve independent Spring Boot services.
Every figure below comes from production systems I built or rebuilt — measured against the baseline that existed before, not against a benchmark I chose afterwards. Percentages are approximate where the original instrumentation was.
500K+
Customers
12
Microservices
−35%
API latency
−30%
Security findings
200+
Retail sites
−45%
Report time
−65%
Search time
−70%
Proposal prep
99%
Classification
10K+
Docs indexed
1,000+
Pages / min
−60%
Manual ops
High-throughput APIs built for resilience first. I move legacy monoliths onto event-driven architectures that handle millions of transactions without losing one.
Containerised, observable, zero-downtime deployment. Automated pipelines so your team ships on a Friday afternoon without flinching.
Retrieval systems and agent workflows that remove real operational hours — not demos. Grounded in your own documents, measured against your own baseline.
When the backend needs a face. Responsive interfaces wired straight into secure APIs, so one person owns the whole vertical slice.
The same four steps whether it is a monolith that needs splitting or a document pile that needs a retrieval layer. No surprises, no black boxes.
Read the system as it actually is — traffic shape, failure modes, compliance constraints, and the deadlines nobody wrote down.
Design the target state and, more importantly, the migration path to it. Every step ships independently and can be rolled back.
Implement with tests, CI/CD, and observability wired in from the first commit — not bolted on the week before launch.
Load test against real traffic patterns, write the runbooks, and transfer it properly so your team owns it after I leave.
Most engagements start with a short call about what is breaking and what it is costing you. No pitch deck — just a technical conversation.
Book a 15-minute callWhat working together looks like
No black boxes, and no surprises in week six.
Every architectural call comes with the trade-off it made and the option it rejected. If you disagree, you can say so before it is built.
Work ships in increments that each move a number you already track. You never have to take my word for whether it is going well.
Runbooks for the failures we actually hit, dashboards someone else understands, and a proper handover. A system only I can debug is not finished.
Xebia · Digital customer onboarding platform
500K+
Customers
−35%
API latency
−30%
Security findings
Uworx / Flooid · Point-of-sale platform
200+
Store locations
−45%
Report time
−25%
Store config time
StackWise · Enterprise document intelligence
<500ms
p95 retrieval
1000+
Pages parsed / hr
100%
Traceable citations
StackWise · Enterprise RAG engine
Deep RAG
Retrieval
100s
Documents indexed
Source
Cited answers
StackWise · RFP response automation
−70%
Prep time
1000+
Pages / min
DOCX
Native output
Steel fabrication & prefab manufacturer, UAE
Live
In production
SEO
Search-first build
Masjid & cultural centre, London
Live
Daily use
Donations
Integrated
School management system, UK
Multi-role
Access control
SaaS
Architecture
BSc Software Engineering — CGPA 3.55 / 4.0
How I think about the problems before the code — architecture calls, trade-offs, and the things that only show up in production.
Open to full-time engineering roles, contract work, and technical consulting — backend architecture, cloud migration, and AI systems integration. I reply to everything within a working day.
Pick a slot that suits you — you’ll get a Google Meet link automatically.