Auto Data Sync
Keep records consistent across the tools a team already uses.
App A↔App B
Hello, I’m
EVANNARAYA
Backend Engineer(AI & AutomationInfrastructure).
APIs · backend systems · AI integrations
HOW I BUILD
I design for
what happens
when systems
fail..
Reliable APIs, explicit state, and recovery paths. Not just the happy path.
See systems and proof →SELECTED WORK
Work you can
inspect.
Lead intake, documents, and controlled AI.
01 / 03 · FULL-STACK ENGINEERING
Full-Stack AI Operations Console
A working Next.js interface, Fastify API and PostgreSQL worker system. Configure workflows, approve execution, inspect saved attempts, and recover failed runs.
58/58 system tests · 10/10 browser cases in each native and Docker environment · 10/10 Docker smoke checks
Deterministic HTTP sandbox for AI triage; no live production LLM or customer deployment.
Next.js · React · TypeScript · Fastify · PostgreSQL · Prisma · Docker
RUNNING APPLICATION / CI CAPTURE
WORKFLOWS · APPROVALS · RECOVERY
PRACTICAL AI IN BUSINESS
Less busywork.
More work done.
What could AI automation actually help your team do? Here are practical examples—not claims about live customer deployments.
01 / SALES OPERATIONS
Auto Lead Follow-Up
A customer asks about a service. The system checks the request, prepares the right next step, and keeps follow-ups from being forgotten.
Explore related workflow code ↗Illustrative example · not a measured client outcomeCustomer inquiriesILLUSTRATION
JD
New service inquiry“Could someone call me about a quote?”
ReceivedCheckedAssignedFollow-up
✓ Next action prepared for the team
02 / DOCUMENTS
Auto Document Processing
Instead of copying details from forms and PDFs by hand, documents are read, organized, and checked before someone uses the information.
Explore document-system proof ↗Illustrative interface · based on a tested engineering referenceDocument inboxILLUSTRATION
FORM / PDF
Information extractedName Sample CustomerReference DEMO-001Status Needs review
✓ Source available for checking
03 / INTERNAL TEAMS
Internal AI Assistant
A teammate asks a question about a document or process. The assistant finds relevant information and shows where the answer came from.
Explore knowledge-system proof ↗Illustrative conversation · not a live workplace assistantTeam knowledgeILLUSTRATION
Where can I find the current onboarding steps?
✳
Here is the relevant section.
Open the onboarding guide and review the checklist for your team.
Source: Onboarding Guide · exampleAuto Reporting
Gather routine updates into a report that people can review.
Autopilot Tasks
Move repeatable tasks forward while sending exceptions to a person.
To do→Review
These are conceptual business examples, not screenshots of client production systems. The linked repositories document related engineering capabilities; individual examples are not claimed as separately delivered customer features.
02 / 03 · PYTHON / FASTAPI
Bounded Agent Runtime
AI actions need boundaries and safe recovery. This reference records intent before API calls, holds sensitive writes for approval, and reconciles uncertain outcomes.
261 tests · 90.01% coverage — preserved v2
Source & tests →ARCHITECTURE / AGENT RUNTIME
EXPLICIT EXECUTION STATESIMPLIFIED SYSTEM MODEL
01 / ADMISSIONValidatetenant + request
02 / CONTROLApprovepersistent decision
03 / WRITEPersist intentbefore external I/O
04 / RECOVERReconcileuncertain outcome
Timeout ≠ failure.Avoid repeating a possible external side effect.
API · STATE · RECOVERY
03 / 03 · FASTAPI / PGVECTOR / MCP
Knowledge Workflow Runtime
Find answers in current documents. This reference processes files, tracks versions, returns source citations, and holds uncertain extraction for review.
134 tests · 91.64% coverage · 12/12 evals
Source & tests →ARCHITECTURE / KNOWLEDGE RUNTIME
SOURCE LIFECYCLESIMPLIFIED SYSTEM MODEL
01 / SOURCEIngestPDF · image · DOCX
02 / STATEVersionatomic updates + deletes
03 / RETRIEVALRetrievecurrent source citations
04 / AUTHORITYReviewMCP + human approval
Stale data stays out.Current citations and review before handoff.
RETRIEVAL · DATA · AUTHORITY
Independent tested reference. Results are archived v2 evidence, not current commercial use or universal exactly-once delivery.
Default retrieval uses deterministic lexical embeddings and extractive answers. Twelve synthetic regression cases do not establish commercial semantic retrieval quality.
These are tested engineering references, not claims of live customer scale or uptime. Test figures describe the linked revisions. Agent evidence · Knowledge verification
ALSO BUILT
More engineering
work.
TYPESCRIPT / POSTGRESQL
Production Integration Reference
Lead intake with signed webhooks, duplicate-event checks, bounded retries, and saved handoff outcomes.
Runnable reference · tests · CI
N8N / JAVASCRIPT
VAREVANT Workflow Engineering
Lead routing and follow-up controls: suppression, state checks, no-send paths, and recovery evidence.
117-node historical graph · 60 Code nodes
Also: BIMMCA Intelligence ↗ — Supabase browser consumer.
WHAT I BUILD
AI can assist.
Rules still decide.
Route leads.
Validate lead intake and API handoffs. PIR source ↗
Find answers.
Find cited answers in current documents. Source ↗
Control AI.
Review sensitive writes; reconcile uncertainty. Source ↗
CAPABILITIES
The stack behind
the work.
Backend
Python · FastAPI
PostgreSQL · Pydantic
Integration
n8n · REST APIs
Webhooks · TypeScript
AI systems
MCP · Retrieval
Structured outputs · OCR
Reliability
Idempotency · Approvals
Tests · CI · Recovery
SELECTED CLIENT WORK
Client-facing work.
PT Geget Gigit · Indonesia
AI-assisted marketing operations automation.
EZUmrah · Malaysia
Six-workflow automation and integration scope for travel operations.
CONTACT
Let’s
connect.
Backend Engineer specializing in Python, FastAPI, Supabase, and AI Automation Workflows.