Eojin isn't taking new orders for this project right now. Here are some similar projects to explore.
You will get a PostgreSQL NL2SQL guardrail audit with failure reproduction
Rising Talent
Project details
AUDIT, NOT A BUILD. I review one existing Node.js/PostgreSQL Text-to-SQL path to identify why it produces unsafe, incorrect, or unstable SQL. Using sanitized files, one failure example, schema constraints, and your read-only policy, I reproduce the issue when possible, map the request path and guardrails, and deliver prioritized findings plus acceptance tests. Boundary Check covers one path, one failure, and the top three risks. Failure & Guardrail Audit adds sanitized reproduction, schema/allowlist/read-only analysis, and a prioritized report. Remediation Test Plan adds evaluation cases, acceptance tests, fix order, and an evidence receipt. This project does not build a new NL2SQL system, change code, access production databases or credentials, deploy changes, or certify security or accuracy.
Database Type
PostgreSQLWhat's included
| Service Tiers |
Starter
$199
|
Standard
$449
|
Advanced
$799
|
|---|---|---|---|
| Delivery Time | 3 days | 5 days | 7 days |
Number of Revisions | 1 | 1 | 2 |
Number of Queries | 1 | 1 | 1 |
Query Debugging | - | - | - |
Query Optimization | - | - | - |
Query Scheduling | - | - | - |
Query Analysis | - | - | - |
Source Code | - | - | - |
Frequently asked questions
About Eojin
AI & Backend Engineer | Node.js, Python, Java, Azure
Yeongdeungpo-gu, South Korea - 6:49 am local time
I have nearly eight years of software engineering experience across enterprise legal AI, SaaS, document intelligence, cloud automation, and IoT/device platforms. My core stack includes Node.js, NestJS, TypeScript, Python, Java, Spring Boot, PostgreSQL, Docker, and Microsoft Azure.
I can help you with:
• AI agent and LLM-backed application architecture
• Provider-agnostic model integrations that make new providers and models easy to add
• REST APIs, backend services, database design, and third-party API integrations
• Customer-isolated Azure environments automated with Bicep and Docker
• OCR/NER-based document processing and PII detection or masking PoCs
Selected experience:
• Built the complete backend for an enterprise legal AI agent SaaS and designed its provider/model adapter layer.
• Automated customer deployments from Azure resource groups through services and containers, removing manual infrastructure setup.
• Rebuilt a legacy operations platform supporting approximately 100,000 taxi and bus terminals, including TCP monitoring, inventory, statistics, remote deployment, and status tracking.
• Built an LwM2M monitoring platform for LTE Cat-M1 modems with registration, health monitoring, remote commands, and firmware updates.
I take ownership from architecture through deployment, communicate tradeoffs clearly, and design systems that are straightforward to extend and operate. If you need an engineer who can turn an AI or backend requirement into a deployable system, let’s discuss your project.
Steps for completing your project
After purchasing the project, send requirements so Eojin can start the project.
Delivery time starts when Eojin receives requirements from you.
Eojin works on your project following the steps below.
Revisions may occur after the delivery date.
Submit sanitized inputs
Submit the sanitized architecture summary, versions, one failure or reliability question, redacted examples, permitted query boundary, acceptance owner, and deterministic acceptance check.
Confirm safety boundaries
Confirm the access, data classification, retention/deletion, and no-secrets boundary. Missing or unsafe requirements are DO_NOT_START.