droideronline

droideronline / pyconf-hyd-2026-trustworthy-llm-agents

Public

Hands-on workshop: building trustworthy LLM agents with observability (OpenTelemetry, Langfuse), evaluations (DeepEval), and security guardrails

14
5
100% credibility
Found Mar 14, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A workshop demo of a multi-agent customer support system with a chat UI, database, and AI agents for handling orders, policies, refunds, and escalations.

How It Works

1
🔍 Discover Support Swarm

You find this cool demo project online or at a tech conference, promising smart AI helpers for customer support.

2
📦 Prepare your setup

Copy a simple settings file and add your AI thinking service details to connect the brains.

3
🚀 Start everything

Run one easy command to launch the full support team, database, and chat screen.

4
🌱 Add sample customers

Quickly fill in fake customers, orders, and help articles so the agents have real data to work with.

5
đź’¬ Chat like a customer

Open the friendly chat window and ask about orders, refunds, or policies—watch the agents route and respond perfectly.

âś… Experience magic support

See refunds processed, emails sent, and tough queries handled confidently, ready to build your own trustworthy AI team.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is pyconf-hyd-2026-trustworthy-llm-agents?

This GitHub repo delivers a hands-on workshop demo for PyConf Hyd 2026, spinning up a multi-agent customer support system with LangGraph and LangChain on Python backend, paired with a TypeScript Next.js chat UI and PostgreSQL/pgvector database. Users get a full-stack app where queries route to specialized agents—like shop assist for orders/refunds or policy advisor for inquiries—complete with semantic search and email tools. Fire it up via Docker Compose for instant testing, including seed data and a prompt injection demo to showcase security guardrails.

Why is it gaining traction?

It stands out as a ready-to-run example blending observability (OpenTelemetry, Langfuse), evaluations (DeepEval), and security in LLM agents, skipping boilerplate setup for quick iteration. Developers dig the single-command Docker launch, live chat at localhost:3000, and real-world workflow like intent routing plus pgvector-powered knowledge base search. The hands-on workshop format makes it a go-to for generative AI tinkering without starting from scratch.

Who should use this?

AI engineers prototyping agentic customer support bots, LangGraph newcomers wanting multi-agent routing examples, or workshop attendees at PyConf Hyd 2026 practicing observability and guardrails. Ideal for backend devs evaluating trustworthy LLM pipelines before productionizing order lookups or escalations.

Verdict

Grab it for learning hands-on trustworthy agents—docs nail quickstart and architecture—but with 13 stars and 1.0% credibility, it's an early workshop prototype, not battle-tested for prod. Solid foundation to fork and extend. (187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.