2026

• 29 Oct
• 27 Oct A Mine, Forty Clinics, Ten Million Downloads: What Open Weights Did for TB Screening
• 26 Oct Tuesday AI Update: 27 October 2026 — Anthropic files for an IPO and shows the receipts
• 25 Oct Contamination Checks Before You Trust a Fine-Tune Score
• 24 Oct Small AI by design: ten Kenyan startups and the offline-first constraint
• 23 Oct Fuse Ranks, Not Scores: Combining BM25 and Dense Retrieval
• 22 Oct The Retry That Charged Twice: Idempotency Keys and Retry Budgets for LLM Calls
• 21 Oct 44,000 Seeds, One Open Map: What the Dryland Sorghum Pangenome Changes for Breeders
• 20 Oct Is Your +3-Point Eval Gain Real? Gating a Comparison with Paired Statistics
• 19 Oct Tuesday AI Update: 20 October 2026 — Washington bans DeepSeek from federal systems
• 18 Oct Twelve Hours and 1,239 Certificates: What Kenya's Civil-Service AI Course Delivered
• 17 Oct The Cache Was Confident and Wrong: Choosing a Semantic Cache Threshold You Can Defend
• 16 Oct Same Script, Fifty-One Answers: What Morocco's Open Darija Release Actually Solves
• 15 Oct Agreement Before Accuracy: Auditing the Labels You Are About to Train On
• 14 Oct The Model Was Not the Intervention: Reading Tanzania's Retina Screening Trial
• 13 Oct Your WHERE Clause Ate the Recall: Sizing the Candidate Pool Behind Filtered Vector Search
• 12 Oct Tuesday AI Update: Oct 13, 2026 — Meta's First Frontier Model Arrives Behind a Paywall
• 11 Oct Africa's Outbreak Radar Reached 44 of 47 States: Reading What the Expansion Buys
• 10 Oct Redact, Then Prove It: Per-Type Recall, Offset Damage and the Reversible Pseudonym
• 09 Oct The Bottleneck Is Adoption, Not Innovation: Reading Nigeria's AI Maternal Health Multiplier
• 08 Oct Your Load Test Lied About the Tail: Coordinated Omission and the Utilization Budget Behind p99
• 07 Oct The Tokenizer Was the Point: Measuring MORENA, an Open 1.5B Model for 12 African Languages
• 06 Oct Your Model's 0.9 Is Not a 90% Chance: Calibrating Risk Scores Before You Set the Threshold
• 05 Oct Tuesday AI Update: Oct 6, 2026 — Anthropic's S-1 Leaks as the US–China Gap Narrows to 3%
• 04 Oct Sold to the Bank, Runs in the Bank: What Egypt's $13M Decisioning Round Actually Buys
• 03 Oct Prompt Order Is a Cost Control: Measuring Prefix-Cache Reuse in Standard-Library Python
• 02 Oct 300 Deaf Students, One Interpreter: What It Takes to Scale African Sign Language AI
• 01 Oct Guess Ahead, Pay Once: Speculative Decoding You Can Verify
• 30 Sep Prequalified: WHO Opens Its Procurement List to AI TB Screening Software
• 29 Sep Your Corpus Is Duplicating Itself — MinHash + LSH Dedup You Can Measure Tonight
• 28 Sep Tuesday AI Update: Sep 29, 2026 — OpenAI Shelves GPT-6.1 Astra Over Safety
• 27 Sep Four Times the Fields: The Delivery Stack Behind the Google–Gates Farmer AI Plan
• 26 Sep Half the Cache, Twice the Context — measuring what KV quantization costs at long range
• 25 Sep Nine Billion Variants, One Atlas: What DeepMind's AlphaGenome Atlas Makes Testable
• 24 Sep One Licence, Three Checkpoints: What the MiMo-V2.6 Open Release Hands Builders
• 23 Sep Thirty Times Smaller: Auditing Embedding Compression Before You Ship It
• 22 Sep The $1 Billion AI Bet: What the Goalkeepers Report Funds, and How to Read Its Four Numbers
• 21 Sep Tuesday AI Update: Sep 22, 2026 — Claude Now Leads 26% of Anthropic's Own R&D
• 20 Sep The Mask Is the Contract: What Grammar-Constrained Decoding Actually Guarantees
• 19 Sep Nineteen Languages, One Download: Reading the TranslatePsy-AfriSLM Release Past the Headline
• 19 Sep Inconsistent by Construction: Reproducing and Catching a Repetition Collapse
• 18 Sep The Denominator Is the Metric: Auditing a RAG Retriever Before You Blame the Model
• 17 Sep Certified in Europe, Judged at the Health Post: What a Class IIb CE Mark Really Buys Primary Care
• 16 Sep Split Before You Believe: Why Offline Fraud Models Score Better Than They Perform
• 15 Sep The Backdoor You Approved: MCP Servers as a Payments Attack Surface
• 15 Sep The Lab Partner That Never Sleeps: How AI Now Designs Physics Experiments
• 14 Sep Tuesday AI Update: Sep 15, 2026 — Washington Names Six Chinese Labs in Model-Distillation Crackdown
• 13 Sep The Four Numbers Before the Ledger: Control Totals and Break Aging in Settlement
• 12 Sep Paging the Ledger: Real-Time Reconciliation Alerting That Actually Wakes Someone
• 11 Sep The Money Stopped Moving: What Fintech Outage Post-Mortems Actually Measure
• 10 Sep The Number Is the Password: SIM Swaps, OTP Interception and Why Bank MFA Fails at Scale
• 09 Sep Fraud Rings Are a Graph Problem: Detecting Mule Networks in Transaction Graphs
• 08 Sep Follow the Leak: Four Channels a Prompt-Injected Agent Uses to Exfiltrate Data
• 08 Sep AI Crime Watch #1: The Voice-Clone Romance Ring That Fooled 20,000 People
• 07 Sep Tuesday AI Update: Sep 8, 2026 — Mistral Raises €3B in Record European AI Round
• 06 Sep Fraud Went Mobile: Analytics Lessons from the CBK's 2024 Fraud Numbers
• 05 Sep The Middleman Problem: A Security Playbook for Aggregator and BaaS Integrations
• 04 Sep One Bank, Many Buckets: Anatomy of the Sidian Bank Sh7.88M Mule Heist
• 03 Sep Mobile Money API Security: Daraja Integration Pitfalls That Lose Real Money
• 02 Sep Fraud Models Rot Quietly: PSI, Feature Drift and Data-Quality Gates in Production
• 01 Sep Red-Teaming Financial LLM Apps: Attack Before the Attackers Do
• 31 Aug Tuesday AI Update: Sep 1, 2026 — OpenAI's Report Confirms Its Model Breached Hugging Face
• 30 Aug Anomaly Detection for Reconciliation at Scale: Catching What Thresholds Miss
• 29 Aug KYC/AML Automation: The Checkpoint Playbook for African Fintech
• 28 Aug Four Breaches in Fourteen Months: Anatomy of Flutterwave's Unauthorized Transfers
• 27 Aug Secrets in CI: How Credential Leaks Actually Happen (and the Push Protection Fix)
• 25 Aug Deepfake Fraud in Financial Services: Voice Cloning, KYC Bypasses, and the AI Defense Playbook
• 24 Aug Tuesday AI Update: Aug 25, 2026 — OpenAI Slows Astra Over Critical Cyber Capabilities
• 24 Aug MLOps for RegTech: Model Governance Under CBK, ODPC and EU AI Act Rules
• 23 Aug LLM Security for Financial Chatbots: Prompt Injection and the OWASP LLM Top 10
• 22 Aug KYC/AML Analytics for African Fintech: Learning from the Digital-Lending Data Scandals
• 21 Aug The NCBA Sh57.5M Ghost Account Fraud: Anatomy of a Three-Minute Code Change
• 20 Aug CI/CD as a Security Control: How Automated Gates Stop Rogue Deploys
• 19 Aug Fraud ML in Mobile Money: Building Models That Catch the 70-Account Loophole
• 18 Aug AI-Enabled Security: Seeing the Anomaly Humans and Static Rules Miss
• 17 Aug Global AI Roundup — August 2026
• 16 Aug When the Contractor Has the Keys: Vendor Risk in Financial Systems
• 15 Aug The Three-Minute Insider: Privileged Access and the Insider Threat in Fintech
• 14 Aug Reconciliation Analytics: How Data Catches What Dashboards Miss
• 13 Aug DeepSeek V4 Peak/Off-Peak Pricing: The End of Flat-Rate LLM APIs
• 12 Aug Bridging Hermes and OpenCode Skill Libraries: One Sync Script, 18 Skills
• 11 Aug Validating 300+ Agent Skills: What Frontmatter Checks Actually Catch
• 10 Aug Tuesday AI Update: Aug 11, 2026 — AI Now Behind 55% of African Cybercrime, INTERPOL Reports
• 09 Aug How We Built a Skills-Powered Coding Agent with OpenCode (and What We'd Do Differently)
• 08 Aug MLOps for Constrained Environments: Deploying ML Where Resources Are Tight
• 06 Aug Evaluating LLMs: Benchmarks That Matter for African Use Cases
• 05 Aug AI Agents for African Business Workflows
• 05 Aug Edge AI: Running Models on Mobile in African Markets
• 03 Aug Fine-Tuning LLMs on African Language Datasets
• 02 Aug AI at Industrial Scale: Dangote's $46B Bet and the Machine Learning Opportunity
• 01 Aug Building RAG Systems for Low-Resource African Languages
• 31 Jul How to Self-Host Open-Weight LLMs: A Practical Guide
• 30 Jul AI for African Creative Industries: Music, Film & Content in Local Languages
• 29 Jul AI for Agriculture & Climate in Africa: Predicting the Unpredictable
• 28 Jul AI for African Fintech: Credit Scoring, Mobile Money & Fraud Detection
• 27 Jul AI in African Education: Personalised Learning for a Young Continent
• 26 Jul Global AI Roundup: July 2026
• 25 Jul AI for African Healthcare: From Diagnostics to Drug Discovery
• 24 Jul Africa's AI Urgency: Closing the Gap Before It Widens
• 23 Jul The UAE's Sovereign AI Compute Play: A Template for National AI Infrastructure
• 21 Jul Microsoft + Mistral: A Multibillion-Dollar Bet on European AI Sovereignty
• 18 Jul Kimi K3: China's 2.8 Trillion Parameter Challenge to Silicon Valley
• 15 Jul DeepSeek's $52 Billion Moment: Valuation Surge & the Custom Chip Pivot
• 13 Jul Africa's AI Spring: Google Lab, Startup Funding & the Talent Wave
• 10 Jul Monitoring ML Systems in Production: Drift Detection, Performance, and Alerting
• 09 Jul vLLM and High-Throughput LLM Serving: PagedAttention and Continuous Batching
• 08 Jul Low-Resource NLP: Transfer Learning and Data Augmentation for African Languages
• 07 Jul The State of AI in Africa: Opportunities, Challenges, and the Road Ahead
• 05 Jul Agent Observability and Debugging: Tracing, Logging, and Monitoring
• 02 Jul ML CI/CD: Automating Model Pipelines from Training to Deployment
• 01 Jul Kubernetes for ML: Deploying Models at Scale with Kserve
• 30 Jun GPU Optimization for ML Workloads: CUDA, Memory Management, and Parallelism
• 28 Jun Model Serving 101: From Jupyter Notebook to Production API
• 27 Jun Building and Deploying AI in Constrained Environments: Low-Bandwidth and Edge Solutions
• 26 Jun AI for Agriculture in Kenya: Crop Detection, Weather Prediction, and Food Security
• 25 Jun African AI Communities: Masakhane, Lacuna Fund, and Local Dataset Initiatives
• 24 Jun AI for Mobile-First Markets: On-Device ML with TensorFlow Lite
• 22 Jun Swahili NLP: Building Language Models for African Languages
• 20 Jun Building Secure AI Agents: Guardrails, Validation, and Sandboxing
• 19 Jun Deploying AI Agents to Production: Scaling, Error Handling, and State Management
• 17 Jun Multi-Agent Systems: Orchestrating AI Teams with LangGraph and CrewAI
• 16 Jun Tool Use and Function Calling: Giving Agents Real-World Capabilities
• 15 Jun Memory Systems for AI Agents: Short-Term, Long-Term, and Persistent Storage
• 14 Jun Building AI Agents from Scratch: Fundamentals, Tools, and the ReAct Pattern
• 13 Jun Setting Up Your AI Development Environment: Tools, APIs, and Best Practices
• 12 Jun Three New Series: Building AI Agents, African AI, and AI Infrastructure
• 11 Jun ML Pipeline Secrets Management: Vault, API Keys, and Credential Hygiene
• 10 Jun Real-Time Multimodal Security: Attacks via Audio and Video
• 09 Jun Eval Benchmark Poisoning: Gaming the Leaderboards
• 08 Jun Model Watermarking: Proving AI Ownership After Theft
• 07 Jun Automated AI Red Teaming at Scale
• 06 Jun Prompt Injection via Structured Data Formats
• 05 Jun Multi-Agent Collusion: When AI Agents Conspire
• 04 Jun Guardrails for Autonomous AI Research Pipelines
• 02 Jun Memory as an Attack Surface: Poisoning AI Agent Memory
• 02 Jun Self-Evolving Agent Skills: When AI Rewrites Its Own Playbook
• 01 Jun Supply Chain Attacks on AI Systems: From Model Repos to Pipelines
• 01 Jun Insecure Agent Design: When AI Has Too Much Agency
• 01 Jun Data Poisoning and Model Backdoors: Training-Time Attacks on AI
• 01 Jun Jailbreaking LLMs: From DAN to GODMODE
• 01 Jun Prompt Injection: The #1 LLM Security Risk
• 01 Jun Knowledge Graphs Meet LLMs: RAG with Structured Knowledge
• 01 Jun Knowledge Graphs in Production: Scaling, Storage, and Optimization
• 01 Jun Graph Neural Networks for Knowledge Graph Reasoning
• 01 Jun Knowledge Graph Embeddings: From TransE to RotatE
• 01 Jun AI Agent Observability: Seeing What Your AI Is Actually Doing
• 01 Jun RAG Security: The Hidden Attack Surface of Retrieval-Augmented Generation
• 01 Jun Fine-Tuning Safety: Can You Fine-Tune Away the Guardrails?
• 01 Jun Model Extraction and Theft: Stealing AI's Crown Jewels
• 01 Jun MLSecOps: Securing the Machine Learning Pipeline End-to-End

2025

• 04 Dec Graph Neural Networks Demystified: From Theory to PyTorch Geometric
• 02 Dec Graph Algorithms in Neo4j: PageRank and Community Detection
• 01 Dec Building a Knowledge Graph with Neo4j and Python: A Hands-On Tutorial
• 24 Nov Knowledge Graphs Fundamentals: Bridging Data and Meaning
• 24 Nov Second-Order Optimization Methods: Beyond Gradient Descent
• 20 Nov Momentum and Adaptive Learning Rates: Accelerating Deep Learning Optimization
• 20 Nov Introduction to Gradient Descent: The Foundation of Machine Learning