Netthplus
The commercial desk that never sleeps
Netthplus is a Colombian IT consultancy — SAP, cybersecurity, cloud, DevOps, RPA and IT talent. We gave them a WhatsApp AI assistant trained on their own technical documents that handles first contact in Spanish, grounds every answer in their real portfolio, qualifies the lead and alerts a human advisor the moment the conversation calls for it.
Duration
Ongoing · live since July 2026
Started
July 8, 2026
End to end in production, including retrieval and four AI detectors
From 11 Netthplus documents, all indexed in pgvector
Model tag, keyword safety net and commitment detector
26 policies; server-side keys never reach the browser
Netthplus has a wide portfolio — SAP FI/CO and S/4HANA, cybersecurity, cloud, DevOps, RPA, AI analytics and IT staffing — and a small sales team.
Every inbound WhatsApp enquiry had to be typed by hand, and nights and weekends quietly lost pipeline to competitors.
- — Most inbound questions were repetitive — services, SAP modules, consultant profiles, certifications — and answered one by one
- — No coverage outside office hours: prospects writing at night or on weekends went elsewhere
- — Scattered history: each advisor kept chats on their own phone, with no consolidated leads and no organizational memory
- — Marketing had no single channel to point campaigns at — profiles and business cards pointed to whoever was on duty
Five systems on the WebMinds WhatsApp AI platform, white-labelled for Netthplus at app.
netthplus.com, with code and domain owned by the client.
- + WhatsApp AI agent · Meta Cloud API webhook with signature verification, Gemini generation grounded in retrieval, voice notes and customer images understood, replies in Colombian Spanish
- + Document knowledge base · PDFs and office files chunked, embedded at 768 dimensions and served from pgvector, with hybrid full-text plus semantic search and vision-based reading of scanned PDFs
- + Human escalation · three independent triggers notify the sales advisor in parallel by an approved WhatsApp template and by email with the latest messages attached
- + Lead extraction and CRM lite · a second AI pass pulls name, email, interest, estimated value and status from each conversation into a lead board
- + Operations dashboard and social entry points · real-time inbox, training studio with chat simulator, analytics, and a wa.me + QR generator per channel so every campaign points to the same assistant
Inside the build.
Brand, interface and system — shown as they shipped.
Not a bot. Commercial infrastructure.
Behind every reply runs a retrieval pipeline, LLM generation, four secondary AI detectors, multichannel notifications and row-level security on every table. The webhook acknowledges Meta in under two seconds and does the heavy work in the background. Roughly 65% of the effort is backend.
Answer Meta first, think later
The webhook returns 200 OK in under two seconds and keeps processing in the background, so slow model calls never trigger Meta retries or leave a prospect waiting.
Survives the next model release
When a newly promoted Flash model saturated for days, a pinned stable version with a cascading fallback kept the assistant answering. Retries per model, then a fallback chain, with every recovery logged.
Room to think
Reasoning tokens count against the output budget, which cut replies mid-word. A larger budget and logging of every non-normal finish solved it for good.
Escalation that survives abandonment
The assistant flags the lead both when it asks for contact details and when it confirms them, with a six-hour cooldown, so a prospect who stops halfway still reaches an advisor.
Compliance the client controls
The data-policy link the bot cites when it collects personal data (Colombian Law 1581) is editable from the panel, so the client can update it without a deployment.
Hybrid AI Coding
Senior WebMinds engineering steers AI agents for implementation, with independent QA on every change. Architecture, product judgment and client conversations stay human; every production finding becomes a documented fix and a reusable diagnostic script.
Technologies & Tools
The tech stack powering this solution
Beyond the numbers, we delivered a system that transformed how this client operates.
From hours to seconds
Enquiries that used to wait in a manual inbox now get an answer in about six seconds on average, around the clock, grounded in Netthplus’s own documents. People step in when the assistant escalates, alerted in parallel by WhatsApp and email.
One channel for all of marketing
wa.me links and downloadable QR codes per channel — Instagram, Facebook, email signatures, business cards — give marketing one assistant to point every campaign at, with a prefilled message that shows where each lead came from.
Knowledge and pipeline, consolidated
Netthplus’s portfolio lives in one searchable knowledge base, and conversations and extracted leads live in one secured database — auditable, exportable and ready for their CRM.
- Mar–Jun 2026Platform foundation: RAG, leads, catalog media, simulator, web widget, social links and hybrid search
- Jul 8, 2026Launch for Netthplus: Meta compliance, Netthplus branding, legal pages and the first real conversation
- Jul 2026Two-turn escalation, keyword safety net and template-language fix after early feedback
- Aug 21, 2026Fast acknowledgement to Meta and a cascading model fallback
- Sep 2026Client-editable data-policy link and a help guide for sharing the assistant on social media
- OngoingContinuous operation and support
Ready to achieve similar results?
Let's discuss how we can transform your business with a tailored digital solution.