Work
Problems, and what
was built for them.
Client systems first, organised by the problem they removed. Products and platforms after, organised by what they do. Every entry opens a write-up with the architecture, the decisions, and what was hard.
Client systems
7 case studies
01Sales Automation
AI Lead Pipeline for a Sales Team
A B2B services team was hand-sorting 200+ inbound leads a week, with reps spending more time on data entry and triage than selling.
80%
less manual sorting
3x
faster first response
02Knowledge Retrieval
Production RAG for Enterprise Documents
An internal team lost hours every week digging through scattered PDFs, wikis, and ERPNext records to answer routine questions.
85%+
retrieval accuracy
Hours
saved per team weekly
03Customer Support
Voice AI for Customer Support
A growing support team was buried under repetitive Tier-1 inquiries, driving long wait times and agent burnout.
60%
resolved without a human
24/7
availability
04Sales Operations
Business Cards to CRM Records, Automatically
A team came back from every trade show with a stack of business cards, and the cards that got typed into the CRM were the ones somebody still had energy for on the flight home.
Same day
card to CRM record
Zero
manual retyping
05Manufacturing & Solar
Custom ERPNext for Manufacturing and Solar Operations
Operations ran on ERPNext for accounting and on spreadsheets for everything the business actually did, so the numbers in the system and the numbers people used were never the same.
One
system of record
0
spreadsheets in the critical path
06Outbound Sales
Outbound That Reads Like It Was Written for One Person
Outbound was a template with a first name in it. Reply rates reflected that, and the fix everyone reaches for, sending more of it, makes the domain worse rather than the pipeline better.
Per lead
copy written from real context
Own domain
sending reputation kept in house
07Marketing & Content
A Content Engine for LinkedIn and Short-Form Video
Content is the one growth channel that stops the week everyone gets busy, because every post is a research task, a writing task, a design task, and a scheduling task pretending to be one job.
Queued
publishing survives a busy week
Variants
produced per campaign, not one draft
Featured systems
Products and platforms
AI Legal Assistant & Contract PlatformLawBot AfricaAI legal platform for African jurisdictions: an assistant that answers in plain language, a contract generator, and a document analyser, in one workspace.Live
Voice AI Platform for Sales & SupportChatterbots.aiPlatform for building AI voice assistants that hold a real conversation, answer from your own content, qualify the caller, and book the meeting.Live
AI Hiring & Voice Screening PlatformMarqHire AIHiring platform that parses resumes, scores them against a job description, and runs first-round screening interviews by voice.Production
AI Website Builder for Small BusinessesBrandryDescribe your business in one sentence and Brandry builds a real, mobile-ready website. No code, no designers, no jargon.Live
Multi-Agent Document IntelligenceAgentFlowMulti-agent platform that turns complex business files into cited, verifiable executive briefs.Live
AI Health & Wellness PlatformWellnixAI health platform combining motion analysis, real-time nutrition intelligence, and predictive wellness.Live
9 more, each with its own write-up.
The archive carries every system, filterable by AI and agents, automation, full stack, or open source, with the status stated on each card. Open any one for the problem, the architecture, and the decisions behind it.