Critical paths
Real money, below the API.
Transaction-critical processing in HLASM and COBOL — described at the level the work can safely be discussed in public.
Joakim Moléni Systems engineer · Stockholm
I write code where money actually moves.
From HLASM and COBOL on IBM z/OS to JavaScript applications on OpenShift. End-to-end production ownership throughout.
Scroll, swipe, use the arrow keys or choose an address below.
0x04 / Context
I work where abstractions run out.
The code beneath the API still decides whether a transaction completes and the overnight run finishes. That is where I work: HLASM and COBOL on IBM z/OS, with production and on-call responsibility.
My career did not begin when LinkedIn started calling me a developer. I had already built web systems, administered Linux, owned businesses, led operations and shipped work with consequences. The title caught up later.
COBOL was the route back in. My first change went to production in the week after training. I joined daytime on-call within three months and moved into first-call night duty after two years. Today I am a leading technical contributor to the move from an assembler-driven platform to modern COBOL while the system remains live.
- PROD
- Code and on-call ownership
- 50+
- Engineers using my tools
- 850+
- Knowledge-platform pages
- 100k
- Fragsheet registrations
Register map / current state
R0SYSTEMS// role
R1HLASM// core.lang
R2COBOL// core.lang
R3z/OS// platform
R4PROD// environment
R5ON-CALL// responsibility
R6LIVE// migration.mode
R750+// tooling.reach
R8850+// docs.shipped
R9100k// product.users
R10OPENSHIFT// runtime
R11CI/DB// delivery.data
R12JS/CSS// applications
R13OPS// history
R14OWN// operating.style
R15READY// status
One operating model across every layer
0x08.A / Core systems
Register-level reasoning.
Transaction-critical workloads in HLASM and COBOL. Diagnostics, controlled change and incident response in systems that do not get to be approximately correct.
- Constraint
- The system stays live.
- Failure mode
- Production gets the call.
- Standard
- Traceable, reviewable, deterministic.
0x08.B / Platforms & products
The whole path, when it matters.
OpenShift applications, JavaScript services, database-backed workflows, CI pipelines, internal developer tools and AWS product delivery. I own the path from architecture and integration through deployment and production operation.
- Built
- Applications, data flows, APIs and delivery pipelines.
- Operated
- OpenShift, Linux, AWS and production services.
- Principle
- Choose tools. Own consequences.
Different runtimes. Same requirement: understand the system and own the consequence.
0x0C / Systems
The systems are old. The problems are not.
Modernization
Replace the engine while it runs.
A leading technical contributor to a controlled move from assembler-driven processing to modern COBOL, stepping into coordination and technical decisions when needed.
Production ownership
The incident does not end at the ticket.
On-call response, cross-system diagnostics, root-cause investigation and controlled change back into live environments.
Modern production systems
Own the whole path.
Integrated internal applications on OpenShift: JavaScript, database-backed workflows, CI pipelines and production integrations used across engineering and leadership.
0x10 / Proof
Proof, under constraints.
The banking source stays inside the bank. The responsibility, constraints and engineering choices can still be described precisely.
Employer work · confidential
Core banking platform migration
A leading technical contributor to the move from assembler-driven processing to modern COBOL, with additional coordination responsibility when needed while the platform remains live.
- Constraint
- Live system
- Judgment
- Controlled change under constraint
Internal system · built
Engineering knowledge platform
850+ pages with custom routing, full-text search, markdown rendering, syntax highlighting and a build pipeline — used daily across the department.
- Constraint
- Replace fragmented knowledge
- Ownership
- Architecture through delivery
Internal applications · production
Integrated operational platforms
Production applications on OpenShift connecting JavaScript services, database-backed workflows and CI pipelines across engineering and leadership processes. Owned end-to-end from architecture and integration through deployment and operation.
- Runtime
- OpenShift / JavaScript / databases
- Ownership
- Architecture through production
Developer tooling · built
Tools for mainframe engineers
AI-assisted review, instructions and architecture guidance designed for the actual mainframe workflow, adopted by 50+ engineers.
- Constraint
- Specialist codebase
- Evidence
- 50+ engineers
LIA internship · product work
Fragsheet
Integrated ML coaching models through product connectors and adapters, rewrote the full web platform in vanilla JavaScript and migrated the infrastructure from DigitalOcean to AWS. 100,000 registered users.
- Range
- PyTorch → web → AWS
- Evidence
- 100k registrations
0x14 / Contact
Send the difficult thing.
I am interested in core systems, live modernization, developer platforms and products that need someone willing to own the whole path.
If the work is mostly rearranging abstractions, I may not be your person. If correctness, migration risk or technical ownership actually matter, we should talk.
01ResumeChoose the relevant emphasis
02Emailjcrooge@gmail.com
03LinkedInConversation and work history
04GitHubPublic code
Available for consequential systems, hard migrations and real ownership.
© 2026 Joakim Moléni