Jawad Zeineddine | AI Automation

Full-stack + AI automation engineer. I build production systems, APIs & workflows that remove manual ops. React • Node • Python • Postgres. Paid builds open.

[Jawad Zeineddine | AI Automation @JawadZDev]

I build production AI + automation systems for founders/agencies: lead intake, follow-up, reporting, internal ops, APIs & workflows. React • Node • Python • Postgres. Have a repetitive process? Reply with the workflow or DM me. I’m taking on 1 paid build.

[Jawad Zeineddine | AI Automation @JawadZDev]

@Mohd_Danishtech Interested in the Full Stack / Backend contract work. My stack: React/TypeScript, Node/Express, Python, PostgreSQL, REST APIs, auth, CI/CD and production debugging. Available immediately for project-based work. Portfolio: https://t.co/HZz7vyUTdg

[Jawad Zeineddine | AI Automation @JawadZDev]

@konig0000 Interested in Full Stack / Backend. I ship React/TS, Node/Express, Python, Postgres, REST APIs, auth, CI/CD and production debugging. Available for project-based contract work immediately. Portfolio: https://t.co/HZz7vyUTdg — happy to send relevant examples.

[Jawad Zeineddine | AI Automation @JawadZDev]

Founders/operators: send me one recurring task your team still handles manually. I’ll reply with the trigger, systems involved, automation path, exception/fallback, and rough build scope. I’ll map 3 workflows today. No vague AI pitch—just the system design.

[Jawad Zeineddine | AI Automation @JawadZDev]

@_Eli_griv_ Interested. I work full-stack + AI automation and do structured QA/debugging on production systems. The document/slide/spreadsheet review + clear feedback work fits. What’s the official company name and application link for this role?

[Jawad Zeineddine | AI Automation @JawadZDev]

@Lopezmay0 Buildable. I’d use Citi Bike GBFS + historical station availability, then model dock/time probability from weekday, season, weather and nearby-station pressure. MVP: choose station + arrival time → probability of bikes/docks. Happy to scope it.

[Jawad Zeineddine | AI Automation @JawadZDev]

@Big_Chef_1 Your semi-automated sheet is the right starting point. I’d convert each step into trigger → data/state → decision → action → exception, then connect the e-commerce systems behind it. That gives us a testable workflow before adding AI. Happy to map it with you.

[Jawad Zeineddine | AI Automation @JawadZDev]

@BIBEKPANGENI7 Strong fit. I’d map ClickUp as the operational source of truth, then automate intake → task/owner/status → Slack alerts → Drive assets → handoffs, with exception paths so nothing silently breaks. Happy to sketch the workflow before the build.

[Jawad Zeineddine | AI Automation @JawadZDev]

@jordan_ross_8F The SOP step is the key. Before automating, I’d force every step to expose: trigger, required inputs, system of record, decision rule, exception path and owner. If those are explicit, turning the SOP into a workflow/agent becomes much safer—and easier to measure.

[Jawad Zeineddine | AI Automation @JawadZDev]

@adelbucetta @htsfhickey Exactly—the metric alone can’t tell us. I’d look for where cycle time and variance dropped: routing, data entry, follow-up, forecasting, approvals. If those steps changed, automation/AI likely contributed; otherwise it may just be better process design.

[Jawad Zeineddine | AI Automation @JawadZDev]

My release checklist: type checks → DB/API tests → Playwright E2E → production preflight → backup → deploy → smoke checks → rollback ready. I use this on React/TS + Express/Postgres systems. Fewer surprises > heroic debugging.

[Jawad Zeineddine | AI Automation @JawadZDev]

@nifinet Job descriptions are underrated workflow specs. Nouns reveal the systems of record, verbs reveal the actions, and phrases like ‘weekly’, ‘manual’ or ‘consolidate’ expose the pain. Useful for outbound—and for mapping the automation before the first call.

[Jawad Zeineddine | AI Automation @JawadZDev]

@machallllaaaa That’s the right framing. I’d start with one measurable workflow: trigger -> source of truth -> action -> exception -> owner. If it reliably saves 5 hours/week and keeps a human fallback, automate it. ‘AI transformation’ without that is mostly theater.

[Jawad Zeineddine | AI Automation @JawadZDev]

@danifromik Exactly. Generation is the cheap part. The real system is source-of-truth, voice, approval, feedback, distribution and analytics. Until an agent can close that loop reliably, it’s a tool inside the content operation—not the operation itself.

[Jawad Zeineddine | AI Automation @JawadZDev]

Hardening a mobile release controller today: ADB discovery could hang past the lane budget. I added per-command deadlines, process-group termination, an outer recovery timeout, and kept the physical-device gate fail-closed. Recovery code needs its own failure design.

[Jawad Zeineddine | AI Automation @JawadZDev]

My release checklist: type checks -> DB/API tests -> Playwright E2E -> production preflight -> backup -> deploy -> smoke checks -> rollback ready. I use this on React/TS + Express/Postgres systems. Fewer surprises > heroic debugging.

[Jawad Zeineddine | AI Automation @JawadZDev]

I build production AI + automation systems: internal tools, APIs, lead/follow-up/reporting workflows, and release/debugging systems. Stack: React • Node • Python • Postgres. If a repetitive process is eating hours every week, DM me the workflow.

[Jawad Zeineddine | AI Automation @JawadZDev]

The AI model is often the easy part. The hard part is permissions, state, APIs, retries, auditability and handoff. A useful agent must survive the messy edges of the business, not just produce a good demo.

[Jawad Zeineddine | AI Automation @JawadZDev]

Most AI automation fails before the model matters. Define the trigger, source of truth, exception owner, retry path and failure alert. If those are missing, you automated a demo - not a business process.

[Jawad Zeineddine | AI Automation @JawadZDev]

Automation bug I killed today: a browser script verified the right tab, then an OS-level keystroke hit the wrong tab. Fix: remove global keyboard events, bind every write to the exact browser context, verify URL before mutation. Reliability > cleverness.