CONFIRMED STRENGTH
Backend foundation
C#/.NET, REST APIs, SQL-backed business systems, integrations, live troubleshooting and delivery from requirement to usable service.
See capability map ↗AI FULL-STACK / AI AGENT ENGINEER
Backend depth.
AI application range.
20+ years of software engineering across C#/.NET, REST APIs, SQL-backed systems, integrations and live-system support. Recently focused on OpenAI-enabled RAG, AI agents, tool workflows and short-cycle delivery from POC to usable service.
POSITIONINGBackend-first full-stack integration with honest frontend and platform boundaries.
20+years engineering
POCto usable service
RAGagents + retrieval
EDGELinux / SSH operation
WHY THIS FITS
A practical bridge between
software depth and AI delivery.
The strongest signal is not a job title. It is the ability to understand an existing system, connect the right services, validate a focused idea and leave behind evidence that another engineer can review.
APPLIED AI
RAG, OpenAI and agents
Retrieval, OCR, embeddings, citations, prompt/tool workflows, structured outputs and repository-aware engineering with human checkpoints.
See selected work ↗FULL-STACK INTEGRATION
Connect the layers
Web application and API delivery across data, service and integration layers. React/Next.js and canvas-specific work are not claimed as established specialties.
Read the boundaries ↗CLOUD / EDGE
Operate what gets built
Cloudflare, Azure, AWS, Docker and MIC-733AO Linux/SSH routines for logs, storage, resource checks and service health.
See Edge evidence ↗SELECTED EVIDENCE
Work that shows
how I build and verify.
Each case keeps its status visible: internal service, development use, POC, permission-dependent integration or evaluation-device operation.
01
RAG / RETRIEVAL
INTERNAL AI SERVICE
Local AI and RAG workflows
A private retrieval workflow built around OCR, embeddings, hybrid search, citations and separated knowledge boundaries.
OCREmbeddingsCitationsHybrid retrieval
View role and evidence＋
RoleDefined the retrieval shape, iterated the workflow with real documents and kept source boundaries visible.
- OCR and embedding pipeline
- Source-aware answers and knowledge boundaries
- Internal service rather than a public product claim
02
ENGINEERING AGENTS
AI-ASSISTED DEVELOPMENT
Code review and API documentation agents
Repository-aware workflows that turn repetitive analysis into structured, reviewable output without removing human engineering judgment.
Prompt / toolStructured outputReview gates
View role and evidence＋
RoleDefined inputs, output formats and review rules, then tested the agents against actual repositories.
- Code review and change guidance
- API documentation from implementation evidence
- Human review retained for architecture and security decisions
03
ENTERPRISE INTEGRATION
TEAMS BOT / OPENAB
AI workplace assistant integration
A workplace assistant connecting UI/Ollama, RAG, enterprise knowledge, OpenAB and Teams Bot components, with permission-dependent tenant integration kept explicit.
Teams BotOpenABKnowledge integration
View role and evidence＋
RoleBuilt and validated the short-cycle foundation, then extended notification audit, Jira handoff deduplication and issue classification.
- Local Teams Bot health and message paths
- OpenAB Gateway integration design
- Tenant credentials and admin approval remain prerequisites
04
EDGE AI / OPERATIONS
MIC-733AO
Linux / SSH service operation
Multiple internal AI and backend services operated on an Advantech MIC-733AO evaluation device with routine checks for logs, storage, resources and service health.
LinuxSSHLogsService health
View role and evidence＋
BoundaryThis is Edge service operation and maintenance context; it is not presented as manufacturing-system ownership or model-training experience.
- SSH-based service checks
- Log, storage and CPU/memory follow-up
- Internal service operation on an evaluation device
05
CLOUD DELIVERY
CLOUDFLARE / EDGE
From zero-to-one POCs to safer delivery
Short-cycle commerce, public-data, education and transit prototypes, with delivery safeguards around source sync, secrets scope, staging, health checks and rollback checks.
CloudflareEdge deliveryHealth checks
View role and evidence＋
RoleHandled product flow, backend integration, delivery checks and validation for focused prototypes.
- Scope-limited credential and token-presence checks
- Source synchronization and staging exclusions
- Service health and rollback follow-up
06
LEGACY MODERNIZATION
PL/SQL + PRO*C TO .NET
AI-assisted API migration workflow
A controlled path from legacy source analysis to .NET implementation planning, with task decomposition, tests, review gates and blocker evidence.
Source analysis.NETReview gates
View role and evidence＋
RoleUsed AI-assisted engineering to break down legacy API work while keeping scope, evidence and human review visible.
- PL/SQL and Pro*C source analysis
- Task decomposition and implementation planning
- Tests, contract checks and review checkpoints
CAPABILITY MAP
A stack for
real integration work.
The list separates established hands-on experience from areas that are intentionally left open for growth.
Layer
Confirmed experience
Positioning
AI application
OpenAI API, RAG, OCR, embeddings, citations, prompt/tool workflows, agents, structured outputs
Applied AI integration, not model training
Backend
C#/.NET, REST APIs, SQL Server, MySQL, SQLite, business systems, live troubleshooting
Primary engineering depth
Cloud / Edge
Azure, AWS, Cloudflare, Docker, IIS, Linux/SSH, MIC-733AO service operation
Delivery and operation context
Integration
Teams Bot, OpenAB Gateway, Confluence/Jira knowledge integration, APIs
Connect services and preserve boundaries
Growth areas
React/Next.js, WebGL/Canvas, Kubernetes, MLOps, model training
Not presented as established skills
RESUME DOWNLOADS
Ready for a closer
look at the evidence?
Two A4 versions are prepared for review. Both keep the same factual background, with the role alignment and language adjusted for the reader.
CONTACT