Real systems. Real results.
25 years enterprise IT. Now building end-to-end AI automation platforms for businesses that want to move faster without hiring more people.
Case studies
Manual B2B prospecting across Australia and Canada was slow, inconsistent, and impossible to scale beyond a few contacts per day.
Fully automated system — scrapes prospects, enriches with AI research, writes personalised WhatsApp messages, qualifies replies, and books calls. Zero human involvement until a meeting is confirmed.
Manually writing and posting LinkedIn content was taking 45–60 minutes every day. Consistency was poor and ideas were running out.
Claude generates posts using a dual-formula system, fal.ai creates matching images, a Gmail approval gate gives human sign-off, then LinkedIn API publishes automatically 4x per week.
HR team manually reviewing 50–100 resumes per role, each taking 5–10 minutes. Backlog causing candidate drop-off and inconsistent scoring decisions.
n8n reads incoming resumes from Gmail, extracts text, sends to OpenAI for scoring against job criteria, logs results in Google Sheets, and auto-sends rejection emails for scores below threshold.
Lead follow-up happening manually via WhatsApp, email, and LinkedIn separately. Leads going cold due to delayed responses with no consistent sequence.
When a lead submits a form, the system sends a WhatsApp acknowledgement, personalised email within 2 minutes, logs in Sheets, and fires a follow-up WhatsApp at 24 hours if no reply.
Creating short-form video content manually required scripting, recording, editing, and uploading — hours per video, impossible to scale to volume.
Daily schedule fires, DeepSeek writes the script, Claude polishes the hook, Kokoro TTS narrates in Hindi, fal.ai generates scene images, FFmpeg stitches the video, YouTube API publishes — fully automated.
Businesses processing invoices, RFPs, and contracts manually — extracting data by hand, filing in spreadsheets, losing hours to repetitive document review.
Documents uploaded, OCR-extracted via Dockling, analysed by Claude for structured data extraction, classified by type, stored in Supabase, and surfaced through an Appsmith dashboard.
Businesses paying $500–2,000/month for SaaS tools that don’t integrate, can’t be customised, and send sensitive data to third-party servers they don’t control.
Complete self-hosted AI automation platform — 38 containerised services on one VPS covering workflow automation, AI agents, vector search, WhatsApp, CRM, social scheduling, voice AI, and a branded client dashboard.
Tech stack
Ready to automate
your business?
Book a free 30-minute AI audit. I’ll map out exactly what can be automated and what it’s worth to your bottom line.