OPERATIONAL SYSTEMS · DESIGNED, BUILT, RUN

I rebuild how companies run.

Operations redesigned as production systems — not decks, not pilots. When the spreadsheets, hand-offs and swivel-chair work finally break, this is who gets called.

LIVE · PRODUCTION TELEMETRY02:14:07
intake.parsenominal
erp.sync0 manual touches
exceptions.queue3 open
invoice.dispatchok
99.9%
UPTIME SLA
80+
PRODUCTION SERVERS
24/7
UNATTENDED
10y+
LONGEST SYSTEM LIVE
25+
YEARS IN PRODUCTION
130+
SYSTEMS SHIPPED
16
INDUSTRIES

Your operation already runs on systems. Most were never designed.

01 · THE PROBLEM
GLUE-01

Glue-work

Your best people spend their days re-typing data between tools that were never introduced to each other. That's not a staffing problem — it's a missing system.

SWVL-02

Swivel-chair operations

Order in one screen, status in a second, invoice in a third. Every swivel is a delay, an error budget, and a person who can never take a holiday.

PLOT-03

Pilot purgatory

The AI demo impressed everyone eight months ago. It still isn't in production, because demos are easy and systems are not.

One method. Four stages. No theater.

02 · THE METHOD
D-01

Diagnose

Two weeks inside your operation. Where the hours go, where the errors breed, what the real constraint is — which is never what the brief said.

Output: written findings,
quantified, with a verdict.
A-02

Architect

The target system on one page: components, flows, owners, failure modes. Boring technology chosen on purpose. Decisions documented with their reasoning.

Output: architecture +
build plan you could shop around.
B-03

Build

Shipped in slices, each one live in production before the next begins. Your team operates slice one while slice two is built. No big-bang cutover, ever.

Output: a running system,
documented and handed over.
R-04

Run

Monitored, maintained, evolved. The system earns its keep quarter after quarter — or you'll see exactly why in the numbers, because everything is measured.

Output: uptime, telemetry,
and a quarterly systems review.

Selected systems

03 · TRANSFORMATIONS, NOT PROJECTS
E-COMMERCE200+ FTE10 YEARS RUNNING

Unifying a 200+ employee e-commerce operation

A growing e-commerce company serving multiple European markets had disconnected systems — NetSuite ERP, PrestaShop webshop, invoicing, payment gateways — each drifting apart daily. I architected the end-to-end event-driven sync, built the finance core, and run the production infrastructure.

READ THE CASE STUDY →
200+
Employees supported
80+
Production servers
99.9%
Uptime SLA
10Y+
Client partnership
PRINT & PACKAGINGGLOBAL900+ INSTALLS

Global prepress automation, 900+ installations

Print and packaging companies worldwide need automated PDF preflight, correction, and production tracking. Deep callas pdfToolbox integrations eliminated manual prepress work across 900+ installations — years of reliable operation without spec-level regressions.

READ THE CASE STUDY →
900+
Installations
10Y
In production
0
Spec regressions
24/7
Automation
IOT & ACCESSFULL STACK14 MONTHS

Smart video doorbell & access control, from scratch

A hardware company needed a complete smart access platform: video doorbell with sub-second calling, NFC-based access, and a management dashboard. Outdoor, all-weather, patent-pending. One engineer owning the full stack — firmware to cloud to apps.

READ THE CASE STUDY →
<200ms
Video latency
14mo
Idea → shipping
99.9%
Unlock success
1
Patent filed
Sinisa Devcic — founder & systems architect, DynConnect
04 · WHO YOU'RE DEALING WITH

You're not hiring an agency. You're hiring the person who builds it.

I've spent twenty-five years inside other people's operations — print floors, warehouses, hardware labs, e-commerce back offices. A hundred and thirty systems shipped to production across sixteen industries. The pattern never changes: the company isn't short on software, it's short on design.

I work alone, with your team. No account managers, no juniors learning on your budget. The person who diagnoses your operation is the person who architects the system, writes the code, and answers the phone when something breaks.

Sinisa Devcic
FOUNDER & SYSTEMS ARCHITECT, DYNCONNECT

Three ways in. One of them costs less than the problem costs you per week.

05 · ENGAGEMENTS
ENTRY · FIXED SCOPE

Systems Diagnostic

€4,500 · 2 WEEKS

Two weeks inside your operation. You get a written, quantified verdict on where the hours and errors actually go — and what to build first.

  • Process + systems inventory
  • Quantified findings report
  • Prioritized build plan
  • Yours to keep, or shop around
Book the diagnostic
CORE · PROJECT

Operational Rebuild

FROM €45,000 · 8–16 WEEKS

The full method: diagnose, architect, build. A production system shipped in slices, documented, and handed over to your team.

  • Architecture + decision record
  • Production system, shipped live
  • Team handover + documentation
  • 30-day post-launch cover
Discuss a rebuild
ONGOING · RETAINER

Run & Evolve

FROM €3,000 / MONTH

Your systems, monitored and evolved. Telemetry reviewed monthly, improvements shipped quarterly, one phone number when it matters.

  • Monitoring + incident response
  • Quarterly systems review
  • Continuous small improvements
  • Priority access for new builds
Talk retainers

Systems Journal

06 · FIELD NOTES
ALL FIELD NOTES →
07 · START

No discovery calls. No proposals. A diagnostic, with a verdict.

Tell me about your operation. If it's a fit, we book the two-week diagnostic. If it isn't, I'll tell you that in the first email — and usually point you somewhere better.

STEP 01

You write

Three sentences about your operation and what's breaking. No forms longer than that deserve to exist.

STEP 02

I answer — with a verdict

Within two working days: fit or no fit, and why. No "let's hop on a call to explore synergies."

STEP 03

We diagnose

Two weeks, fixed price, written findings. Everything after that is your decision, made with real information.