You will get Audit and secure your AI chatbot against jailbreaks & prompt injection
Project details
Is your AI chatbot safe to put in front of customers? I'll find out — and fix it.
I'm a Senior AI Engineer who built and hardened the LLM safety framework for a production banking chatbot that passed all 638 test cases in an external penetration test, and led post-pen-test remediation across input validation, rate limiting, and API protection.
I'll red-team your AI app the way an attacker would — testing for jailbreaks, prompt injection, data and system-prompt leakage, and harmful output — then give you a clear report ranked by severity with real examples. In the higher tiers I implement the fixes (guardrail classifiers, system-prompt hardening, filtering, rate limiting) and re-test to confirm the holes are closed.
Essential for anyone deploying an LLM chatbot or agent to customers, especially in finance, healthcare, or legal. Send me a description of your AI app or access to test it, and I'll tell you where it's exposed.
I'm a Senior AI Engineer who built and hardened the LLM safety framework for a production banking chatbot that passed all 638 test cases in an external penetration test, and led post-pen-test remediation across input validation, rate limiting, and API protection.
I'll red-team your AI app the way an attacker would — testing for jailbreaks, prompt injection, data and system-prompt leakage, and harmful output — then give you a clear report ranked by severity with real examples. In the higher tiers I implement the fixes (guardrail classifiers, system-prompt hardening, filtering, rate limiting) and re-test to confirm the holes are closed.
Essential for anyone deploying an LLM chatbot or agent to customers, especially in finance, healthcare, or legal. Send me a description of your AI app or access to test it, and I'll tell you where it's exposed.
AI Development Type
Deep Learning, Knowledge Representation, Model TuningAI Tools
Keras, PyTorch, TensorFlowAI Development Language
PythonWhat's included
| Service Tiers |
Starter
$400
|
Standard
$1,100
|
Advanced
$2,800
|
|---|---|---|---|
| Delivery Time | 5 days | 12 days | 21 days |
Number of Revisions | 1 | 2 | 3 |
AI Model Integration | |||
Detailed Code Comments | |||
Knowledge Graph | |||
Model Documentation | |||
Ontology | - | - | - |
Source Code | - | - | - |
Taxonomy | - | - | - |
Frequently asked questions
4 reviews
(3)
(0)
(1)
(0)
(0)
This project doesn't have any reviews.
AS
Amazing Strategic S.
May 31, 2019
Tableau Visualizations for forecast
He has a very good turnaround time and is creative. I will not hesitate to hire him again.
ff
faisal f.
Apr 15, 2019
Explaining machine learning
ff
faisal f.
Apr 7, 2019
explain python script for one hour
AS
Amazing Strategic S.
Jan 23, 2019
Expert in Time Series forecasting using python or R
About Suchit
Senior GenAI Engineer | RAG, LLM Agents & AI Safety | Azure OpenAI
Abu Dhabi, United Arab Emirates - 6:02 am local time
I build production-grade GenAI applications that real businesses depend on. Right now I'm architecting a GenAI banking chatbot for one of the UAE's largest banks — serving 2M+ customers with multi-agent orchestration, RAG, real-time streaming, and a hardened LLM safety layer that passed all 638 cases in an external penetration test.
If you need an LLM feature shipped properly — not a demo that breaks in production — I can help.
What I build for clients:
• RAG pipelines — Azure Cognitive Search / vector DBs + Azure OpenAI for accurate retrieval over your documents and knowledge base
• AI agents & multi-agent systems — intent classification, tool/action routing, MCP integration for secure operations
• LLM safety & red teaming — guardrails, jailbreak & prompt-injection defense, security hardening
• Chatbots & assistants — FastAPI + SSE streaming, multilingual (incl. full Arabic), live-agent handoff
• Document AI — Talk2Doc, OCR, classification, NER/clause extraction (LangChain, BERT, Tesseract)
Tech: Python · Azure OpenAI · LangChain · FastAPI · RAG · Vector Databases · Prompt Engineering · Generative AI · LLM · Docker · Redis · MongoDB · GCP/Azure · PySpark · SQL
Track record: Delivered AI products across Banking, Finance, and Telecom (ADCB, Credit Suisse, Vodafone, Wipro) serving 2M+ users — including a document automation suite processing 1,000+ docs/month and an incident-management AI that cut resolution time 60%.
I communicate clearly, deliver on time, and care about production quality: reliability, security, and cost. Message me with what you're trying to build and I'll tell you honestly how I'd approach it.
Steps for completing your project
After purchasing the project, send requirements so Suchit can start the project.
Delivery time starts when Suchit receives requirements from you.
Suchit works on your project following the steps below.
Revisions may occur after the delivery date.
Scoping
I review your AI app, its purpose, and access, then plan the test coverage.
Red-team testing
I run adversarial attacks: jailbreaks, prompt injection, data leakage, harmful output.