BACKEND ENGINEER · ADDIS ABABA · REMOTE

I build backend systems that hold together as products get complicated.

APIs, data models, workflows, integrations, infrastructure operations, and runtime evidence—from development through production.

View selected work Discuss a backend role
01

WHY ISRAEL

Backend ownership when the system crosses boundaries.

CROSSING BOUNDARIESAPI + RULESDATA + WORKFLOWEXTERNAL SYSTEMSRUNTIME BEHAVIORISRAEL OWNSthe whole flowDEV → PRODUCTIONCOHERENT BACKENDCLEAR AUTHORITYLEGAL TRANSITIONSINTEGRATION CONTROLFAILURE CONTEXT
02

PROOF

Systems that had to hold up after release.

Product backends, developer tooling, infrastructure control, operational workflows, and a payment ledger with measurable impact.

BROWSER STATEBrowser stateCached reflectionCREDIT LEDGERBackend creditledgerSource of truthPOSTGRESQLSTRIPECheckoutWEBHOOKSEventsBALANCE RULESControlledmutationsCONTENT OWNERSHIPAccessenforced
01

NITESHIFT SYSTEMS · SPRING BOOT · STRIPE

I stopped the browser from deciding how much credit a user had.

The old flow trusted browser state, so balances and access logic could drift.

Owned
The Spring Boot ledger, mutation rules, balance checks, and Stripe-linked service flows.
Decision
Put credit authority in one backend model across purchases, subscriptions, and content ownership.
Result
One system of record and fewer ambiguous failure paths.
  • CREDIT LEDGER
  • STRIPE CHECKOUT
  • WEBHOOKS

Credit-related support tickets dropped to zero after release.

EVENT INGESTIONRequest startAPI operationFASTAPIFLOW RECONSTRUCTIONDB queryDB EVIDENCEError foundSOURCE INTELLIGENCECASE FILESCase fileTRACE READY
02

RUNTIME DIAGNOSTICS · FASTAPI · POSTGRESQL

Threadline reconstructs what a backend request actually did.

Scattered logs make it hard to explain a failure as one causal flow.

Owned
FastAPI ingestion, raw event storage, reconstruction, authentication, and source intelligence.
Decision
Build deterministic operation trees and case files from ordered runtime events.
Result
An inspectable request flow linked to database activity, errors, and source context.
  • EVENT INGESTION
  • FLOW RECONSTRUCTION
  • DB EVIDENCE
SPRING BOOTControl planeDEPLOYMENT STATEOWNERENCRYPTED SSHPython layerREMOTE HOSTLinux host 01REMOTE HOSTLinux host 02REMOTE HOSTLinux host 03DOCKERODOOPOSTGRES
03

INFRASTRUCTURE AUTOMATION · SPRING BOOT · PYTHON

One control plane for Odoo instances spread across remote servers.

Loose scripts and terminal history did not provide a durable operational record.

Owned
Host registration, provisioning, encrypted SSH, Docker services, and instance lifecycle operations.
Decision
Keep system state in Spring Boot and execute remote work through a Python provisioner.
Result
One stateful path for provisioning, starting, stopping, and updating Odoo instances.
  • REMOTE HOSTS
  • ENCRYPTED SSH
  • DOCKER
APPROVALSSubmittedREQUESTReviewedCHECKApprovedRULEPaymentPROCESSEDHistoryRECORDEDMASTERCARDWALLET FLOWSDOCUMENT VERIFICATIONSMSDATA MIGRATION
04

ETTA SOLUTIONS · PAYMENTS · OPERATIONS

Business workflows with rules, history, and external dependencies.

Manual and fragmented processes had to become software used in active operations.

Owned
Tender, transfer, payment, verification, SMS, and CNET migration workflows.
Decision
Model approvals, status, retries, records, and hash-based tamper checks inside the backend flow.
Result
Operational processes with explicit history and integration boundaries.
  • MASTERCARD
  • WALLET FLOWS
  • APPROVALS
DEVICEZKTecoINGESTIONNormalizeattendance dataHRMSPayroll APIOPSClient useDEBUGDEPLOYFIELD OPS
05

AHAZ SOFTWARE SOLUTIONS · HRMS · INTEGRATIONS

Backend workflows that connected people, devices, and payroll.

The useful system had to work beyond isolated endpoints and across a daily operational process.

Owned
HRMS and payroll APIs, attendance ingestion, normalization, deployment, and debugging.
Decision
Automate ZKTeco device data into the wider attendance and payroll workflow.
Result
End-to-end support for attendance, payroll, site allocation, and field operations.
  • HRMS
  • PAYROLL
  • ATTENDANCE
02

PROOF

Five systems. One end-to-end view.

Different stacks and domains, with the same responsibility for understanding behavior across boundaries.

01Credit Ledger

NITESHIFT SYSTEMS · SPRING BOOT · STRIPE

CREDIT LEDGERSTRIPE CHECKOUTWEBHOOKS
02Threadline

RUNTIME DIAGNOSTICS · FASTAPI · POSTGRESQL

EVENT INGESTIONFLOW RECONSTRUCTIONDB EVIDENCE
03Odoo Control Plane

INFRASTRUCTURE AUTOMATION · SPRING BOOT · PYTHON

REMOTE HOSTSENCRYPTED SSHDOCKER
04Business Systems

ETTA SOLUTIONS · PAYMENTS · OPERATIONS

MASTERCARDWALLET FLOWSAPPROVALS
05Early Engineering

AHAZ SOFTWARE SOLUTIONS · HRMS · INTEGRATIONS

HRMSPAYROLLATTENDANCE
03

ABOUT ISRAEL

I work best where backend decisions shape the whole product.

I’m a backend engineer based in Addis Ababa, Ethiopia.

My work has covered product backends, internal business systems, payments, infrastructure automation, external integrations, developer tooling, and production support.

The common thread is not one industry or framework. It is taking systems with real rules and operational consequences and making their boundaries, behavior, and responsibilities understandable.

I contribute most strongly beneath the interface: service design, APIs, data models, workflow logic, integrations, and the operational paths that carry a change into production.

BSC COMPUTER SCIENCE · HILCOE · 2022

04

HOW I WORK

A practical method for complex backend work.

AUTHORITYWho decides?STATEWhat is true?FAILURE PATHWhat can go wrong?EVIDENCEHow do we know?Design thereal system
  1. 01

    Trace the full flow. Start at the request and follow rules, data, dependencies, and operational actions.

  2. 02

    Assign authority. Decide which boundary can make each decision and which records are durable.

  3. 03

    Model transitions and failures. Make retries, invalid actions, and recovery paths part of the design.

  4. 04

    Integrate deliberately. Keep external services and infrastructure actions behind clear contracts.

  5. 05

    Leave a runtime trail. Capture enough context to explain what happened after release.

05

EXPERIENCE

Backend work across products, operations, and production support.

01
FEB 2024 — PRESENTREMOTE

NiteShift Systems

JAVA BACKEND ENGINEER

Owned backend workflows across credits, payments, subscriptions, purchases, content access, authentication, admin operations, and production support.

The largest change was replacing frontend-authoritative credit logic with a backend-controlled ledger and synchronization model.

02
JUN 2023 — SEP 2024ADDIS ABABA

ETTA Solutions

SOFTWARE ENGINEER · CONTRACT CONSULTANT

Built systems for procurement, transfers, document verification, payments, messaging, and internal operations.

Integrated Mastercard, wallets, SMS, and PostgreSQL workflows, and migrated legacy CNET data while supporting active operations.

03
FEB 2021 — DEC 2022ADDIS ABABA

Ahaz Software Solutions

SOFTWARE ENGINEER

Built HRMS, payroll, attendance, and field-operation APIs and normalized operational data.

Automated ZKTeco biometric ingestion and supported client systems through deployment, debugging, and production use.

06

BACKEND RANGE

A broad backend range, organized around complete systems.

The stack changes by problem. The work still has to connect application behavior, data, integrations, infrastructure, and runtime support.

Backend applications and APIs

Java, Spring Boot, Python, FastAPI, Node.js, REST APIs, authentication, authorization, and service boundaries.

Data, state, and business rules

PostgreSQL, SQL, Hibernate/JPA, MySQL, MongoDB, relational modeling, transactions, ledgers, and operational history.

Integrations and distributed workflows

External APIs, webhooks, SMS services, biometric devices, Kafka, asynchronous workflows, Stripe, Mastercard, wallets, and legacy migration.

Platform, infrastructure, and runtime

Docker, SSH, remote provisioning, deployment support, unit and integration testing, production debugging, and runtime diagnostics.

Frameworks are tools. The deeper value is being able to follow a backend change through every layer it affects.

07

ROLE FIT

Where this experience maps to backend roles.

These are the backend paths most directly supported by the work above.

Backend application engineering

Product APIs, service boundaries, business rules, and data models.

Java and Spring systems

Spring Boot services, control planes, ledgers, and operational workflows.

Python, FastAPI, and developer tooling

Runtime ingestion, reconstruction, provisioning, and diagnostic tools.

Integrations and workflows

External services, webhooks, devices, approvals, retries, and migrations.

Platform and infrastructure automation

Remote hosts, Docker services, encrypted SSH, and lifecycle operations.

Runtime diagnostics

Event capture, database activity, errors, and source-linked case files.

See the supporting systems

WORK WITH ISRAEL

Need someone who can own the backend flow, not just close isolated tickets?

I’m open to remote backend roles and selected contract work.

The strongest fit is a team dealing with complex product behavior, integrations, internal workflows, developer tooling, infrastructure operations, or production reliability.

AVAILABLE FOR REMOTE BACKEND WORK

Discuss a backend roleView LinkedIn Download resume