Overview
See more here: https://ayenticapp.com
Ayentic is a symbiotic AI platform for customer-facing teams: one hub for WhatsApp, email, and Instagram where AI suggests and humans decide. Instead of delegating the entire first line to a generic chatbot.
I contribute as software engineer and architect on the monorepo: Angular 21 frontend, Node/Express services, PostgreSQL, real-time gateway (Socket.IO, Baileys for WhatsApp), and deployment on Railway with Supabase. The product sits in the Sureva venture umbrella alongside sovereign on-prem consulting; public positioning lives at the Ayentic landing (migrating to ayenticapp.com).
The problem we solve
Organizations drown in fragmented channels. Traditional chatbots optimize for deflection; Ayentic optimizes for context: shared knowledge, burst detection, conversation flagging, and configurable flows so AI assists without erasing human judgment.
Architecture highlights
- Split runtime: webapp, realtime gateway, and worker, so WhatsApp sessions and heavy jobs do not block the UI.
- Stack: Angular 21 (signals, PrimeNG), Node/Express API, PostgreSQL migrations-first, Baileys for WhatsApp, HubSpot forms/CRM integration where clients need it.
- Deploy discipline: TEST before PROD, env alignment (Supabase, Railway), E2E, conventional commits documented in repo skills and ADRs.
- Deployment options: cloud or on-premise copy (nDSG, HIPAA mentioned on marketing) for teams that cannot send data to US hyperscalers.
Product / GTM angle
- Tagline (public): symbiotic AI for support teams, more responses, less chaos.
- Differentiator: not "AI-first deflection" but human + AI collaboration with learnable style per organization.
- Sectors: configurable for municipalities, SMEs, and service teams (pricing and sector blocks on the live landing).
- Team: Switzerland (sales, legal anchor) + distributed engineering, continuous coverage without sacrificing local trust.
What I would highlight
- Concrete engineering decisions (split architecture, WhatsApp constraints, migration-first DB, and product creation and distribution).
- Link to public landing page: www.ayenticapp.com
- No duplicate of marketing copy. Case study tone: implementation, trade-offs, outcomes.
- Using AI agents, skills, harnesses and MCP to make it possible for 1 developer to maintain and develop the complete app