NDC Manchester 2025
2025-12-01..2025-12-04
Manchester, United Kingdom
as the organisers put it
NDC Manchester is a 4-day event for Software Developers focusing on AI & Security, 1-4 December 2025
who it's for
software developers, security engineers, machine learning engineers, platform engineers, devops engineers, engineering managers, .NET developers
topics
MCP and ACP agent protocols; agent memory management patterns; prompt hardening for code generation; compile-time taint analysis; supply chain vulnerability defenses; .NET supply chain protection; cloud misconfiguration remediation; secure GitHub Actions pipelines; designing AI data architectures; secure media provenance with C2PA; observability with OpenTelemetry and OpenLit; AI bias mitigation strategies
| title | author | length |
|---|---|---|
| Agents That Talk Back: MCP, ACP, A2A, Oh my! | Taswar Bhatti | 50m |
Demos show a Python FastAPI MCP server exposing add/multiply/subtract/divide tools, an MCP inspector that executes those tools, and clients using standard IO… | ||
| The question is not when to start threat modeling. It's when to stop | Georges Bolssens | 50m |
Claim: threat models never finish, so stop when they become sufficient not complete. Practical stop rules include timeboxing (e.g. two hours per sprint)… | ||
| Advanced Autonomous AI Agents: Memory, Collaboration and Evolution in Action | Andreas Wänqvist | 58m |
Builds .NET agents with Microsoft Agent Framework wired to models deployed in Azure Foundry, plus evals, tracing and an MCP-based tool server; shares a… | ||
| How to Break AI Systems (Before Someone Else Does) | Gary Lopez, Amanda Minnich | 01h 01m |
Shows real exploit chains: hidden prompts in PDFs and emails that exfiltrate password‑reset URLs, multi‑turn “crescendo” jailbreaks that reliably break many… | ||
| Trust No Input: Taint Analysis at Compile Time | Matteo Di Pirro | 52m |
Claims a compiler can prevent unsafe data flows by encoding taint into types and blocking untrusted values at compile time. Implements a Scala library with… | ||
| Prompt Hardening - Secure Code Generation Using AI | Sean Sinclair | 58m |
A systematic evaluation uses Promptu and an LLM as a judge to run 3,000 code-generation tests across OpenAI models (GPT‑4o mini, GPT‑5 mini with reasoning… | ||
| Splitting the Email Atom: Exploiting Parsers to Bypass Access Controls | Gareth Heyes | |
Shows how RFC-compliant addresses are weaponized to bypass domain checks: encoded-word and Unicode overflows let the Ruby mail library (v2.8.x, 58M+ downloads)… | ||
| Practical Strategies for Successful AI Adoption in your Organisation | Robert Hogg | 48m |
Claims AI typically delivers 20% productivity gains but succeeds only with data quality, governance, logging and guardrails; recommends concrete standards like… | ||
| Exploiting the supply chain | Niall Merrigan | 01h 01m |
AI accelerates classic supply‑chain and social‑engineering attacks: vendor email compromise nets ~$2.7B across ~22,000 incidents; npm/PostHog and SolarWinds… | ||
| Are ‘Friends’ Electric?: What It Means to Be Human Now and Tomorrow in the Age of AI | Lianne Potter | 41m |
AI companions like Replica evolve from grief-driven experiments into subscription products that learn intimate details, lack strong security, and monetize… | ||
| From Ancient Greeks to Modern Geeks - Basic Machine Learning Algorithms | Simon Painter | 45m |
Covers Naive Bayes, linear and logistic regression, k‑nearest neighbours and decision trees with concrete examples and code on GitHub, implemented in Excel… | ||
| Cloud’s Dirty Little Secret: It Was Misconfigs All Along | Karl Ots | 57m |
Most cloud breaches trace to misconfigurations, not exotic exploits: hundreds of Azure reviews show recurring failures like public storage buckets, legacy… | ||
| Skill Degradation: An Empirical Analysis of 400+ AI‑Generated Security Fixes | Pedram Hayati | 41m |
An experiment analyzes 420 submissions from a secure‑coding contest and finds AI patches are faster but shallower: 420 submissions, AI fixes 3–4× larger diffs… | ||
| STOIC Security: Shielding Your Generative AI App from the Five Deadly Risks | Jeff Watkins | 58m |
Introduces the STOIC framework — Stolen, Tricked, Obstructed, Infected, Compromised — and cites concrete attacks: model extraction via millions of API queries… | ||
| Supercharging DevOps with MCP (Without Opening a Security Hole) | Alex Shershebnev | 44m |
MCP (Model Context Protocol) lets LLMs call real tools via a client/server JSON-RPC pattern so models can query Grafana, trigger CI/CD, read Sentry, create… | ||
| Beyond the Pink: Data Risks in Femtech You Can’t Ignore | Arohi Naik | 59m |
Many popular period and fertility apps share precise location, mood, medication and breastfeeding logs with analytics and ad services and sometimes with… | ||
| The EU CRA Readiness for Contributors and Maintainers : A Survival Toolkit | Roman Zhukov | 53m |
EU Cyber Resilience Act is already in force (Dec 2024) with September 2026 reporting for vulnerabilities and full obligations from December 2027; free open… | ||
| OpenSSF Best Practices for Everyone! | Avishay Balter | 37m |
OpenSSF Best Practices Working Group offers concrete tooling and metrics: Scorecard (0–10) runs ~18 checks and is used via CLI, Docker or GitHub Action to flag… | ||
| Deep dive into data streaming security | Olena Kutsenko | 56m |
Real incidents include a health provider leaking 14 million messages via an unauthenticated Kafka broker and a food delivery cluster exposing ~20,000 customer… | ||
| Beyond the origin - secure media provenance with C2PA | Charlie Halford | 55m |
C2PA defines an embedded manifest format that hashes media bytes, records ingredient chains, attaches signatures and optional timestamping and OCSP stapling… | ||
| Observing AI Applications with OpenLit and OpenTelemetry | Carly Richmond | 01h 02m |
Demo uses OpenLit on top of OpenTelemetry to instrument a TypeScript Next.js travel chatbot, sending traces, logs and metrics to a collector and Elastic… | ||
| Why Auto-Fixing Vulnerable Code Needs More Than Good Intentions | Spyros Gasteratos | 56m |
Built an agentic auto-remediation pipeline that initially zero-shot and n-shot prompts hallucinate (including deleting functions or suggesting 200-line… | ||
| Prompting LLMs to use Tools and Resources with MCP | Richard Brough | 12m |
Shows that small open models can be coaxed to call MCP tools and external APIs by embedding concise JSON/YAML schemas and Llama-style instruction templates… | ||
| Breaking and Protecting GitHub IssueOps Pipelines | Alexander Barabanov | 14m |
Shows how IssueOps comment- or label-triggered GitHub Actions can let untrusted forks run privileged jobs and exfiltrate secrets or abuse paid APIs… | ||
| Designing Scalable, Secure Data Architectures for AI at Enterprise Scale | Chisom Nwokwu | 08m |
Companies discover that traditional data architectures break under AI work loads: a launched recommender called Luminina AI spikes sales but shows wrong… | ||
| Your AI Is Still Biased (Even After You Checked) | Neda Maria Kaizumi | 47m |
Claim: bias re-emerges after deployment whenever models are retrained, personalized, or exposed to skewed users; an experimental avatar that reads calendar and… | ||
| ToolShell, Patch Bypass, and the AI That Might Have Seen It Coming | Pedram Hayati, Soroush Dalili | 47m |
A SharePoint chain (ToolShell CVE-2025-49704/49706) yields unauthenticated remote code execution by combining an authentication bypass to toolpane.aspx with an… | ||
| Building Community-Driven Security Analysis for Your .NET Software Supply Chain | Niels Tanis | 52m |
Explains concrete tools and results: an open-source CLI called Fennec Labs that runs OSS Scorecards on solution dependencies, flags transitive CVEs (now… | ||
| 10 tips to level up your ai-assisted coding | Aleksander Stensby | 01h 03m |
Rapid-fire checklist of actionable techniques: use sub agents with separate context to parallelize work (he demos splitting 100 PDFs into concurrent agents)… | ||
| Designing an IAM Strategy for your Engineering Platform | Henry Been | 50m |
Design maps a product + role to an Entra ID Access Package that adds users to intent-named groups, which are used in target systems (Azure subscriptions, Azure… | ||
| .NET supply chain: Protecting against hidden threats | Tom van den Berg | 57m |
Modern .NET projects gain concrete defenses: NuGet audit (now reporting transitive vulnerabilities by default in .NET 10) and net list package reveal… | ||
| GitHub Actions Security: From CI Nightmare to Supply Chain Sentinel | Niek Palm | 52m |
Shows how issue text or PR branches can inject shells and exfiltrate secrets (demo steals secrets from an issue), and counts from the recent worm: ~25,000… | ||
| nothing matches | ||