You will get a deep forensic security audit for your AI or LLM application
Project details
Building an AI-native app? Your current scanner is likely missing 90% of critical vulnerabilities.
Traditional SAST tools are blind to modern LLM risks like Prompt Injection, Excessive Agency, and Vector DB SQLi. They look for "bad strings," missing semantic logic.
I am an AI Security Architect. I built RepoInspect, a deterministic AI engine that merges AST-aware taint tracking with autonomous AI agents. This maps the structural logic of your code and verifies vulnerabilities with zero false positives.
Recent Findings (71 high-severity bugs found):
Dify: 28 SQLi in Vector DB adapters.
Mem0: 23 High Risks (Reranker Hijacking).
OpenAI SDK: 10 High Risks (Command Injection). Etc,
What You Get: A forensic audit of your repository ensuring your LLM workflows, Vector DBs, and Agentic logic are secure. You receive a detailed PDF report mapping vulnerabilities to OWASP (LLM Top 10), complete with mitigation strategies and compliance impacts (SOC 2/GDPR).
Don't leave your AI logic exposed. Let's secure your repo today.
Traditional SAST tools are blind to modern LLM risks like Prompt Injection, Excessive Agency, and Vector DB SQLi. They look for "bad strings," missing semantic logic.
I am an AI Security Architect. I built RepoInspect, a deterministic AI engine that merges AST-aware taint tracking with autonomous AI agents. This maps the structural logic of your code and verifies vulnerabilities with zero false positives.
Recent Findings (71 high-severity bugs found):
Dify: 28 SQLi in Vector DB adapters.
Mem0: 23 High Risks (Reranker Hijacking).
OpenAI SDK: 10 High Risks (Command Injection). Etc,
What You Get: A forensic audit of your repository ensuring your LLM workflows, Vector DBs, and Agentic logic are secure. You receive a detailed PDF report mapping vulnerabilities to OWASP (LLM Top 10), complete with mitigation strategies and compliance impacts (SOC 2/GDPR).
Don't leave your AI logic exposed. Let's secure your repo today.
Cybersecurity Expertise
AI Compliance, Data Protection, Gap AnalysisTechnology Type
IaaS, Database, SaaS, Web Application, PaaSCybersecurity Regulation
GDPR, NIST Cybersecurity Framework, SOC 2What's included
| Service Tiers |
Starter
$350
|
Standard
$950
|
Advanced
$2,500
|
|---|---|---|---|
| Delivery Time | 2 days | 4 days | 7 days |
Small Company Size | |||
Medium Company Size | - | ||
Large Company Size | - | - |
Optional add-ons
You can add these on the next page.
Fast Delivery
+$100 - $500
Post-Patch Verification Scan
(+ 2 Days)
+$350Frequently asked questions
2 reviews
(2)
(0)
(0)
(0)
(0)
This project doesn't have any reviews.
JM
Jean M.
Oct 1, 2021
Data migration testing to validate old data is exactly the same as new data
Ritesh is great to work with. I will definitely contact him again when i have more work.
PO
Pyayt Phyo O.
Jul 10, 2021
Looking for a Knime expert
About Ritesh
AI/ML Expert | LLM & RAG Apps | Python & Streamlit | Databricks | NLP
Kolkata, India - 5:49 pm local time
I specialize in:
🔍 Retrieval-Augmented Generation (RAG): LangChain, LangGraph, LlamaIndex
🤖 LLM Integrations: OpenAI, Gemini, fine-tuning with QLoRA, PEFT
📊 Time Series Forecasting using Spark
🧠 NLP & CV Projects: Summarization, sentiment analysis, OCR
⚙️ Automation & SaaS: Streamlit dashboards, YouTube summarizers, trading apps
🏗️ Data Engineering: PySpark, Databricks, distributed workflows
✨ I also run AI Vision Academy, an initiative that teaches cutting-edge AI through webinars and hands-on projects—so I know how to communicate complex ideas simply and clearly.
🔧 Let’s build smart solutions—fast, scalable, and production-ready.
🛠️ Skills
Python, Streamlit, FastAPI
PySpark, Databricks, SQL
LangChain, LangGraph
LlamaIndex, RAG pipelines
Time Series, Forecasting
NLP, OCR, Text Summarization
RabbitMQ, Async Workflows
Git, VS Code, Azure
Steps for completing your project
After purchasing the project, send requirements so Ritesh can start the project.
Delivery time starts when Ritesh receives requirements from you.
Ritesh works on your project following the steps below.
Revisions may occur after the delivery date.
AST Taint Tracking Description
I will run static analysis to map the "skeleton" of your code, tracing how user input flows into your LLM prompts, Vector DBs, and Agent tools.
Agentic Verification
I will use autonomous AI agents to review the identified vulnerabilities, eliminating false positives by verifying if the threat is actually exploitable.