Skip to content

Practical engineering, data and AI.

Mission-critical systems in production, Cloud FinOps governance and scalable ROI.

Working from strategy to execution to scale digital platforms, lower multi-cloud spend (Azure, GCP, AWS), stabilize critical databases, and ship governed AI in production.

Focus areas

Specialties that connect in practice.

I work across infrastructure, data and technical leadership. The mix depends on the problem, the team and the stage of the company.

Work in this area

Practical automation and agents in the real world.

AI agents and automation pipelines integrated with your company’s systems and data. Practical intelligence applied to daily operations, with secure architecture and measurable productivity.

AI AgentsLLMsRAGAutomation

Career track record

Real results from my trajectory. What changed in operations.

Mission-critical systems and high-impact engineering initiatives directly architected or led throughout my technical and executive career.

Absolute financial figures and internal details are withheld under confidentiality. Every metric below reflects real outcomes from my executive career.

Multi-unit food service & retail

-50%in cloud costs

Half the cloud bill across Azure and GCP.

Multi-cloud architecture review, per-application cost allocation, and transactional database right-sizing.

Combined cloud spend (baseline = 100)

Before
100
After
50
Engineering decisions+
Cost governance:No per-app visibility Allocated per app & squad

Context and constraints

National multi-unit food service and retail group running distributed workloads on Azure and GCP without unit-cost visibility per product or squad.

Alternatives and decision

Reviewed architecture across both clouds, enforced per-application cost tagging, tuned SQL queries, right-sized databases, and embedded FinOps governance into the engineering standard.

Baseline and measurement

Sustained 50% reduction in combined cloud spend against the prior baseline while preserving availability and peak scalability.

Transferable lesson

Cloud spend only drops sustainably when every application has clear ownership and cost efficiency lives inside the engineering routine.

High-scale digital & delivery platform

30xscale across digital delivery

30x delivery scale with a 4.8-rated app.

Mobile and backend re-architecture cutting resource use and latency by 70%, paired with mission-critical Protheus ERP tax integration.

API latency & compute consumption

Before
100% (tightly coupled)
After
-70% latency/compute
Engineering decisions+
Customer rating & ERP:Peak-hour bottlenecks 4.8 App + Universo TOTVS Showcase

Context and constraints

Rapid expansion across digital ordering and delivery channels demanded absorbing steep mealtime traffic spikes without app degradation or tax-issuance failures in the ERP.

Alternatives and decision

Re-architected mobile and backend services to decouple critical paths, introduced strategic caching and leaner payloads, and engineered resilient fiscal integration with TOTVS Protheus ERP.

Baseline and measurement

70% drop in compute consumption and latency, enabling 30x growth in delivery volume, a 4.8 app store rating, and national recognition as a featured case at Universo TOTVS.

Transferable lesson

True scale comes from clean domain decoupling, lean contracts, and end-to-end resilience between customer channels and core ERP finance.

Platform engineering & multi-squad leadership

5 mindeployment window, down from 2 hours

Deployments cut from 2 hours to 5 minutes.

Automated CI/CD with Azure DevOps and ArgoCD, safe rollback, and technical governance across 20+ engineers in 3 squads.

End-to-end production deployment time

Before
2 hours
After
5 minutes
Engineering decisions+
Engineering organization:Manual releases & silos 20+ engineers across 3 squads

Context and constraints

Manual release steps consumed two hours per deployment window, increasing production risk and slowing product delivery across squads.

Alternatives and decision

Standardized CI/CD pipelines on Azure DevOps with GitOps via ArgoCD, automated 1-click rollbacks, clear review rituals, and authored the Engineering Handbook aligning 3 product squads (20+ engineers).

Baseline and measurement

Deployment window reduced from 2 hours to 5 minutes with full auditability and safe daily release autonomy.

Transferable lesson

Sustainable engineering speed comes from deterministic pipelines, smaller batch sizes, and shared architectural standards.

Just-in-Time manufacturing & critical databases

-30%in inventory & operational costs

30% reduction in operational costs & inventory.

Shop-floor Just-in-Time control systems paired with high-availability administration across mission-critical ERP databases.

Operational cost & work-in-process inventory

Before
100% (static batches)
After
70% (-30% cost/stock)
Engineering decisions+
Transactional reliability:Closing-window outage risk HA SQL & Linux clusters

Context and constraints

Excess buffer inventory and delayed shop-floor reporting tied up working capital, while core ERP operations required uninterrupted transactional database uptime.

Alternatives and decision

Built industrial control and Just-in-Time logistics synchronization software tied directly to production lines, alongside high-availability clustering, replication, and SQL tuning across SQL Server, PostgreSQL, Oracle, and Linux.

Baseline and measurement

Verified 30% reduction in manufacturing inventory and operational costs while sustaining continuous fiscal, accounting, and financial ERP workloads.

Transferable lesson

When accurate shop-floor data reaches a resilient transactional database on time, software directly unlocks working capital in physical operations.

Recent articles

Lessons straight from the field.

Project decisions, operational problems and what I learned while solving them.

View all articles
Recent publications58 published articles

Cesar Zanis · Technical Reference & Open Studies

Good decisions have numbers, deadlines, and owners.

This space brings together primary-source studies, open-formula simulators, verified production dossiers, and architecture guides. Peer technical exchanges, architecture, speaking invitations, or complex engineering challenges happen naturally through affinity and referral.

Critical systems architecture, FinOps, Data & AIOpen-formula interactive simulatorsPrimary sources (World Bank, IMF, DORA, IBGE)

Open technical and editorial reference · Direct peer-to-peer contact, zero sales pitch.