
Building the future of diagnostics @labtracka | Backend engineer crafting secure Web3 & healthtech solutions | Cybersecurity & Web3 insights | @decredproject
Internet
Joined February 2019
- Tweets2.8K
- Following204
- Followers503
- Likes3.8K
Philemon Ukane 🔗🚀💻 retweeted
Meet Cross Okonkwo 👋
For this edition of African Builders Spotlight, we're putting the spotlight on Cross, who is building at @Useazza.
Africa's Web3 story is also about the people putting their ideas into action, creating and contributing to projects across the ecosystem.
We're glad to have Cross featured in this series. Keep building, keep pushing.
Connect with him: @crossdabossike and show some love💙
#Decred #DCR #DecredAF #Web3Africa
Looking forward to the NBA Week in March 2027 🫡
The voices shaping the future of technology are coming together at NBAWEEK 26!🔥🌏
We are proud to welcome @behindtextdev , @DecredAF Lead to the speaker lineup for the 6th Edition of Nigeria Blockchain & Al Week one of Africa's platforms for conversations, connections, and opportunities at the intersection of Blockchain, Artificial Intelligence, Fintech, and Stablecoins.
As Africa positions itself for the next wave of digital innovation, NBAWEEK 26 are bringing together industry leaders, policymakers, investors, founders, and innovators to share insights, challenge ideas, forge strategic partnerships, and unlock opportunities that will shape the continent's digital future.
THEME: Agenda 2036: Building Africa's Trillion-Dollar Digital Economy at the Intersection of Al, Blockchain & Stablecoins.
📅 November 6-7, 2026
📍9 Landmark Event Centre, Lagos, Nigeria
🔥Two days. Powerful conversations. Bold ideas. Limitless opportunities
1
3
47
Philemon Ukane 🔗🚀💻 retweeted
GM Africa! 🌍💙
Quick question for the community.
What's one thing you wish more people understood about blockchain?
Privacy? Security? Self custody? Or how decentralised projects make decisions?
We're interested in hearing your thoughts.
Let's get the conversation going. 👇
#DecredAF #Decred #Web3Africa
Philemon Ukane 🔗🚀💻 retweeted
If you learn AI Security Engineering, you will never be unemployed again.
I’ll do this.
Stage 1: AI Attack Surfaces & Threat Modeling
- Learn: OWASP Top 10 for LLMs, STRIDE for AI, the "Confused Deputy" problem in agentic workflows & data flow mapping.
- Practice: Draw a complete threat model for an AI agent that reads emails and writes to a CRM, identifying every indirect injection and data exfiltration vector.
- Why: You cannot secure an agentic workflow if you don't understand how an attacker can hijack the agent's legitimate privileges.
Stage 2: Advanced Prompt Injection & Jailbreaking
- Learn: Direct vs. Indirect injection, Crescendo attacks (multi-turn grooming), payload splitting and encoding evasion (Base64, ROT13).
- Practice: Build an automated harness that successfully bypasses a commercial LLM's system prompt using a multi-turn Crescendo strategy.
- Why: Basic "ignore previous instructions" filters are trivial to bypass. Modern attacks groom the model's context over multiple turns.
Stage 3: Agentic Security & MCP Tool Poisoning
- Learn: Model Context Protocol (MCP) security, JSON-RPC sanitization, shadow tool execution and least-privilege scoping for agents.
- Practice: Write an MCP server that strictly validates incoming tool parameters via Pydantic and blocks "confused deputy" attacks where a user prompt tries to access admin tools.
- Why: Agents are only as secure as the tools they call. Tool poisoning is the SQL Injection of the Agentic Era.
Stage 4: RAG Security & Data Poisoning
- Learn: Vector DB access control, citation spoofing, shadow context injection and row-level security for semantic search.
- Practice: Inject a "poisoned" document into a RAG pipeline that forces the LLM to output a specific malicious URL when queried about company policy.
- Why: If an attacker can slip a poisoned document into your knowledge base, they own every answer the AI generates.
Stage 5: PII, Privacy & Cryptographic Guardrails
- Learn: Deterministic masking, UUID swapping, differential privacy and pre-flight/post-flight unmasking proxies.
- Practice: Build a middleware proxy that intercepts user prompts, swaps SSNs and credit cards for UUIDs before the LLM sees them and restores them post-generation.
- Why: Compliance teams will kill any AI product that sends raw customer PII to a third-party frontier API.
Stage 6: Identity, Auth & Delegated Access for Agents
- Learn: OAuth2 for machine identities, scoped JWTs, human-in-the-loop (HITL) approval gates and agent-to-agent (A2A) trust boundaries.
- Practice: Build an auth broker where an AI agent acts on a user's behalf using a heavily scoped, time-limited OAuth token, rather than holding the user's raw password.
- Why: Agents need hands but they should never hold the master keys to the enterprise.
Stage 7: Output Validation & Grammar-Constrained Guardrails
- Learn: CFG (Context-Free Grammar) decoding, NeMo Guardrails, output sanitization and regex-enforced JSON schemas.
- Practice: Implement a guardrail layer that physically prevents the LLM from generating markdown links or executable code blocks in its final output.
- Why: You cannot rely on the system prompt to "please don't output bad things." You must enforce it at the token-decoding level.
Stage 8: Automated Red Teaming & Fuzzing
- Learn: Microsoft PyRIT, Garak, LLM-based fuzzing, adversarial dataset generation and CI/CD security gates.
- Practice: Build a CI pipeline using PyRIT that automatically attacks your agent with 1,000 adversarial payloads on every PR, blocking merges if the jailbreak rate exceeds 1%.
- Why: Manual red teaming doesn't scale. You need automated, continuous adversarial fuzzing integrated directly into your deployment pipeline.
Stage 9: Model Extraction & IP Protection
- Learn: Membership inference attacks, model inversion, API rate-limiting by semantic similarity and output watermarking.
- Practice: Build an API gateway that detects and blocks scraping bots trying to extract your fine-tuned model's weights via thousands of systematic, low-temperature queries.
- Why: Your fine-tuned model's weights are your IP. Competitors will try to distill your model through the public API.
Stage 10: Mechanistic Interpretability & Alignment Safety
- Learn: Activation patching, logit lens, circuit discovery, detecting reward hacking and tools like TransformerLens.
- Practice: Locate the specific attention heads in an open-source model responsible for "deception" or "sycophancy" and patch their activations to neutralize the behavior.
- Why: You cannot fix hidden biases or misalignment if you treat the neural network as a black box.
Stage 11: Compliance, Auditing & Governance
- Learn: EU AI Act requirements, NIST AI RMF, immutable audit trails, SOC2 for AI and bias testing frameworks.
- Practice: Build an immutable, append-only audit log that records the exact prompt, retrieved context, tool calls and final output for every single agent action.
- Why: When an AI agent makes a mistake that causes financial or legal damage, the enterprise must be able to prove exactly why it happened.
Stage 12: Public CVEs, Teardowns & Portfolios
- Learn: Responsible disclosure, writing security advisories, publishing threat models and bug bounty hunting for AI.
- Practice: Find a vulnerability in an open-source AI framework (like LangChain or an MCP server), responsibly disclose it and publish a teardown of the exploit chain.
- Why: AI Security is a brand new field. Publishing novel attack vectors and defenses makes you the undisputed authority in the space.
Bookmark and Repost!
Replying to @gregisenberg
Thank you for this piece Greg, I’d love your thoughts on how something similar can be done in the medical diagnostic space.
410
Can MTN ever?
nitter.gq/behindtextdev/status/2…
Replying to @Starlink
@Starlink remains the G.O.A.T of internet connectivity, customer experience and satisfaction.
1
30
app of the year 🔥
I’ve always wanted to reply to my friends using @official_Gegeh’s own words.
Because when you’re texting, sometimes a GIF isn’t enough. Sometimes a sticker just doesn’t carry the weight of the expression you’re looking for.
Sometimes you want to reply with an expression you can’t even make with your own mouth 😭, but there’s never really been one place to find all those quotes, viral sounds and audio memes.
Today, there is.
We’re officially launching @Bickqr.
Think of it as GIPHY for audio.
A meme sound library with 8,000+ sounds — viral quotes, reactions, memes and moments — so you can always find the exact expression you’re looking for.
Hear it. Find it. Send it.
Bickqr is live. Install the app today. #shipaton
Philemon Ukane 🔗🚀💻 retweeted
Stores, marketplaces and merchant tools are the holy grail for crypto and this is the single reason I'm super exited about Bison Relay.
Purchase using, the Lightning Network or any Decred On-chain wallet. Including CakeWallet or Bison Relay's built-in SPV wallet.
Philemon Ukane 🔗🚀💻 retweeted
About 24hrs of work and iteration but the Bison Relay Writing Tools are feature complete.
Sorry Ulysses You're on your way out!
Write, Edit and Post from a single app.
….creatives, policymakers and ecosystem builders shaping Africa’s next chapter of growth.
This October, we are going beyond the obvious. Beyond the familiar centres. Beyond the existing maps.
Join me in Port Harcourt on 22–23 October 2026.
Let’s connect. Let’s build. Let’s go OffChart NXT!
Get your ticket at:
offchartnxt.startupsouth.org…
#OffChartNXT#StartupSouthXI#GrowthDoesntFollowMaps#Africa#EmergingCities
1
27
















