Building AI Systems
That Transform Businesses

Challenge — Shield Insurance needed an accurate, explainable way to price health premiums across diverse customer profiles without manual underwriting bottlenecks.

Solution — Built and deployed a regression-based ML system predicting premiums from age, BMI, smoking status and medical history, served through an interactive Streamlit app for underwriters.

97%+ prediction accuracy · underwriting time reduced dramatically · production-ready API.

Challenge — The team was losing leads after-hours and burning time answering repetitive product, pricing and order questions on Instagram and WhatsApp.

Solution — Designed a 24/7 GPT-powered ordering assistant with menu awareness, order capture and human handoff — wired into the team's CRM via n8n.

After-hours conversion recovered · 80% faster response · automated order pipeline.

Challenge — Kwikportt needed to qualify inbound logistics requests instantly and route hot leads to the sales team with full context.

Solution — Built a conversational AI concierge that qualifies requests, captures structured data and pushes enriched leads into the CRM with Slack alerts.

Zero missed inquiries · pre-qualified leads · sales team focused only on hot deals.

Challenge — Operations teams were juggling chat, calendar and spreadsheets manually — losing time on context switching every day.

Solution — Deployed an AI agent in n8n with memory, tool access and Google Calendar + Sheets integrations to schedule, log and update operational data conversationally.

Single chat interface for ops · scheduling and logging fully automated.

Challenge — Leadership needed real-time visibility across sales, support and operations without waiting on weekly reports.

Solution — Built an AI-augmented dashboard layer that pulls from CRM, support and ops tools, with natural-language querying and anomaly alerts.

Reporting time cut by 45% · real-time KPIs · executive natural-language Q&A.