Software Engineering
36.7538° N003.0588° EALT / 0013
00
Operating record / independently verifiable
The record behind
the work.
Practice / scope
Senior Systems Builder for AI agents, SaaS, automation and complex software.
Formal foundation, public performance records and production experience. The important claims remain one click from their source.Top performer on Upwork
Upwork Job Success
Meta · IBM · HackerRank · Stanford
01
Selected work, public contributions and production delivery
Evidence,
not theatre.
Different relationships. One standard: understand the constraint, own the difficult part and leave the system stronger.
Live homepage capture · July 2026Rahva RaamatSelected production infrastructure
Leading estonia eCommerce ecosystem
Scoped AI infrastructure and retrieval architecture for product discovery across a large retail catalogue.
Selected production infrastructure · public catalogue context
36.7538° N / SYS-02
ExpensifyCompleted Upwork contract
Expensify - AI fintech 140+ million $ ARR
A focused engineering engagement completed with a five-star public Upwork record.
Public Upwork work-history evidence
Official documentation capture · July 2026OpenAI Agents SDKOpen-source contribution
Official OpenAI Multi-agent orchestration SDK
Two merged public contributions improving tool naming and multi-agent realtime dispatch behavior.
Public contribution · directly verifiable
Official product visual · July 2026OPI / TaskuTarkSelected production system
Leading estonian EdTech Learning platform system
A multi-role learning platform connecting administrators, teachers, students and parents through coordinated access, content and operational workflows.
Selected production system · public summary
Live site capture · July 2026Evidence-led AI services marketplace
A public marketplace and agent-facing registry for discovering skills, prompts, workflows and agent control planes through inspectable evidence rather than opaque recommendations.
Public product · live implementation
Live site capture · July 2026Governed multi-agent operations platform
A unified platform for configuring, grounding, deploying and monitoring specialist AI agents across customer, revenue, voice and background workflows.
Public product · public feature summary
Public case study noteStanford University AI Systems Architecture
Selected production software delivery for Stanford University, focused on structured sentiment, aspect extraction, target grounding, and abstention controls.
Informed by production software for Stanford University · public note context
36.7538° N / SYS-08
Hermes Agent Platform - N#1 self improving AI agent
Foundational architecture for multi-user gateway sessions in the Hermes Agent framework, introducing deterministic runtime peer mapping and strict user-level cache isolation.
Public open-source contribution · directly verifiable
02
Capabilities / the working set
Systems with
consequence.
Architecture is not a diagram. It is the set of decisions that determines whether the next change takes a day—or a quarter.
I
Practice 1 / 3
AI & agentic systems
Intelligence that can use tools, respect boundaries and improve under observation.
- Production assistants
- Agent workflows
- RAG & retrieval
- Evaluation loops
- Tool orchestration
- Cost and latency control
II
Practice 2 / 3
SaaS & platform architecture
Software foundations that make reliability and future change less expensive.
- Platform modernization
- Multi-tenant systems
- Complex integrations
- Reliability engineering
- Technical audits
- Production hardening
III
Practice 3 / 3
Automation & business systems
Operational systems that connect software, data and human judgment.
- Workflow automation
- Cross-system integration
- Internal tools
- Data synchronization
- Human-in-the-loop design
- Custom platforms
AI
AI integration / practical systems for real operations
For business leaders, operators and enterprise teams
Make AI useful
to the business.
I design and build AI around a concrete business outcome: faster customer response, less manual work, better access to company knowledge or a stronger digital product. It connects to the software you already use, with people kept in control where judgment matters.
Serve customers faster
AI receptionists and service agents can answer common questions, qualify enquiries, book appointments, update your CRM and hand complex cases to the right person.
Business valueMore enquiries handled, fewer missed opportunities and a consistent customer experience.
Turn knowledge into answers
Private AI assistants can search approved documents, product data and operational systems, then give staff or customers a clear answer with the source attached.
Business valueLess time spent searching, faster decisions and knowledge that scales beyond a few key people.
Remove operational drag
Workflow agents can read requests, prepare documents, move information between systems and complete routine steps—with approval before consequential actions.
Business valueFewer manual handoffs, shorter cycle times and more capacity for high-value work.
Where AI tends to pay offStart where work is slow,
Start where work is slow,
repetitive or easy to miss.
- Customers wait because calls, messages or requests queue up.
- Staff repeatedly search, summarize, classify or re-enter the same information.
- Work stalls between inboxes, spreadsheets, CRM, ERP or internal tools.
- An AI pilot shows promise, but cannot yet use real business data or operate safely.
A practical boundary
Good AI integration does not mean automating every decision. It means giving AI a defined job, the right information and clear limits—while people retain control of sensitive or high-impact decisions.
03
Best fit / where senior ownership compounds
The right problem
changes everything.
Established SaaS, e-commerce, education and service businesses dealing with fragmented software, stalled platforms, unreliable integrations or AI initiatives that need senior production ownership.
- A system already exists but has become difficult to change.
- Several tools need to behave like one operation.
- AI must work inside real business constraints.
- The project needs architecture and execution from the same senior person.
- Speculative prototypes without an operating need.
- Commodity landing pages or basic websites.
- Undefined products with no decision-maker.
- Engagements based primarily on the lowest hourly rate.
04
Engagements / clear ways to begin
Start with a
defined move.
Each engagement has a clear decision boundary, useful output and direct senior ownership.
Systems Diagnostic
A fixed-scope examination of architecture, workflows, bottlenecks and failure risks.
OutputFindings, priorities and an actionable system plan.
Build or Modernize
Architecture and production implementation for an AI system, SaaS platform, integration or operational workflow.
OutputAn operational system delivered through defined milestones.
Embedded Technical Leadership
Senior architecture and execution ownership for teams that need direction without immediately hiring a full-time technical leader.
OutputDecisions, implementation leadership, review and stabilization.
05
Method / operating doctrine
Movement,
under control.
The work moves through three chambers. Each one removes a different kind of risk.
Clarity
Find the actual constraint before writing the expensive answer.
- Map the current system
- Identify leverage
- Define measurable success
Architecture
Design boundaries that make reliability and future change affordable.
- Reduce unnecessary complexity
- Plan failure modes
- Protect the next change
Execution
Ship, instrument and stabilize the system in the real environment.
- Production-quality delivery
- Observation and feedback
- Useful knowledge transfer
06
About / the accountable human
Oussama Zeddam
The engineer
behind MAV.
I started coding in 2008 and began working professionally in software engineering in 2013. My formal foundation includes a master’s degree in Software Engineering from Arizona State University. Since then, I have built across full-stack products, SaaS platforms, AI systems and the operational infrastructure around them.
Today I work remotely with international companies that need a senior builder to make a difficult system legible—and then ship it.
Alongside client systems, I founded and lead Algeria's two largest digital ecosystems: DZ Combinator (160,000+ members) for startups, and the Algeria Freelancers Community (130,000+ members)—uniting 290,000+ founders, engineers, and digital builders across the region.
Arizona State University
Master’s degree — Software Engineering
Formal training in software architecture, engineering discipline and the design of maintainable systems.
Verified credential
Verified profile portrait / 2026RoleSenior Systems Builder
BaseRemote
EcosystemFounder · 290k+
RangeAI · SaaS · Automation
ModeDirect senior ownership
07
Contact / qualify the system
Coordinates established
Bring me the system
that is slowing you down.
Describe what is fragmented, unreliable or difficult to scale. I will tell you whether I am the right person to rebuild it.
- Direct senior review
- Clear fit assessment
- No sales handoff
Choose the next step