The Secure Agent Build: From TDD to EDD with Enkrypt AI
In traditional software, we rely on TDD (Test-Driven Development) to ensure code does exactly what we tell it to do. But in the world of non-deterministic AI, unit tests aren't enough. We are shifting the paradigm to EDD (Evaluation-Driven Development).
Most developers build AI agents and hope they don't hallucinate or leak data. In this build-focused afternoon, we are moving from "vibe-checking" to rigorous engineering. Using Enkrypt AI’s enterprise-grade security suite and Qdrant’s high-performance vector database, we will build, stress-test, and harden autonomous agents.
This is a hands-on environment where you make your AI "Audit-Ready" by applying EDD principles to the TDD logic you already know.
What You’ll Build
The EDD Pipeline: Move beyond simple prompts to a metric-based development cycle using LangGraph and Qdrant.
Agentic RAG Infrastructure: Setup a Qdrant vector store to manage agent memory and external knowledge securely.
Automated Red Teaming: Use Enkrypt’s API to launch 20+ attack vectors (prompt injection, PII leaks, and Vector Database Poisoning) against your agent.
Sentry Guardrails: Integrate real-time middleware to block adversarial attacks before they hit your vector store or reach the user.
Agentic Governance: Define and enforce "Golden Rules" using the Enkrypt Policy Engine.
The Agenda
02:00 PM – 02:45 PM: The EDD Manifesto. Why security is the new TDD and a deep dive into the Enkrypt Stack.
02:45 PM – 04:30 PM: The Red-Team Sprint. Build a LangGraph-based agent and watch it fail under automated stress testing.
04:30 PM – 05:30 PM: Hardening the Agent. Implementing Sentry Guardrails and re-running evaluations to prove resilience.
05:30 PM – 06:00 PM: Certified Secure Demos. Presenting "Audit-Ready" reports and networking.
Speakers
Divyanshu Singh: ML Research Engineer (Founding Engineer), Enkrypt AI
Nitin Birur: AI Security Researcher, Enkrypt AI
Who Should Join?
AI Architects & Engineers building production-grade RAG or Agentic systems.
CTOs & Product Leads concerned about AI safety, privacy, and compliance.
Requirements
Laptop with Python environment set up.
Basic familiarity with AI Agents or RAG Pipelines.
An eagerness to break (and then fix) things.
Hosted by
Deepak Chawla – AI Entrepreneur, Founder of HiDevs
Deepak is on a mission to build the world’s largest Gen AI workforce. He has trained 5000+ learners, hosted 100+ AI workshops, and is actively building real-world applications with LLMs, LangChain, and vector databases.
Community Partners
The HSR Founders Club is an exclusive community uniting visionaries and entrepreneurs in Bengaluru. They don't just connect founders; they help them discover the tools, resources, and people to supercharge their journey. They keep meetups small, connections close, and dosa recommendations strong.
Location is AI House