Modernizing Your HCL Connections Experience Four practical, implementable scenarios that transform your Connections deployment — from org charts and custom profiles to sovereign AI and modern APIs. 

Meet Our Expert Speakers Lukas Bühler Full-Stack Web Developer Belsoft Collaboration AG Wannes Rams Cloud Architect ISW 

The Problem with Out-of- the-Box Connections Generic UI Buried Data Zero AI Outdated API's 

The Opportunity Is Right Here Modern Frameworks Angular, React, D3 and other web components can integrate seamlessly with Connections extensibility hooks. Your Data Is Ready The data already exists — profiles, org structures, documents. It just needs a better interface. Sovereign AI On-premises AI is real, capable, and ready — no cloud dependency required. New APIs HCL is actively modernizing the API layer — making Connections more developer-friendly than ever. 

What We're Covering Today Interactive Org Chart Lukas Custom Profile Experience Wannes AI-Powered Integration Lukas Modern Blogs API Wannes 

01 — ORG CHART Interactive Org Chart 

The Idea: Make What's Already There Visible HCL Connections already stores report-to relationships in every user profile, but is a bit dated. The Goal: Transform that existing structured data into a modern, interactive org chart — with Connections as the source of truth. 

The Result: What We Built Interactive Click any person to explore their team, navigate up the reporting chain, or drill into direct reports. Searchable Find anyone in the organization instantly — no more scrolling through directory lists. Hierarchy-Aware Full reporting chains visible at a glance — from individual contributor all the way to the C-suite. Embedded Lives inside Connections via the Customizer framework — users never leave the platform. 

Live Demo — Org Chart Interactive Hierarchy View Report-to Chain with one click Links to Profile, Mail, Telephone Grouped by Country Vimeo OrgChart Demo Engage 26 

TECHNICAL ARCHITECTURE Technical Architecture How It's Built Data: Sourced from HCL Connections Profiles API Backend: Generating organizational chart hirarchy with Caching Layer Frontend: Angular web component passes data to D3 for visual rendering No core product changes — 100% via supported extension mechanisms. Interactive Frontend Angular frontnend with D3 rendered displaying the hirarchy Backend with Caching Layer Generate Organizational Chart with Caching Layer Connections API Connections Profiles endpoint with Report-to Chain data 

"Your org data is already there. We just gave it a face." No New Data Uses report-to relationships that already exist in your Connections profiles. Zero Core Impact Pure extension — no modifications to HCL Connections itself. Instant Value Immediate value for employees, HR, and leadership. 

02 — PROFILES Tailored Profile Experience 

The Problem: Profiles Nobody Trusts Generic by Design Default profile fields were built to be universal — they don't know about your cost centers, skills taxonomies, or department codes. Data Trapped Elsewhere HR data lives in SAP, HR database, or LDAP — and never makes it into Connections profiles. Adoption Killer Users click a profile, find nothing useful, and stop coming back. Profiles become a dead corner of the platform. 

Out of the box 

Custom 

Possible 

How to Feed Custom Data Into Profiles Three Complementary Approaches d TDI Automated Sync — Connect to virtually any HR system and sync data on a schedule. Set it up once; profiles stay current automatically. e User Self-Population — Users can update their personal information using the built in "Edit Profile" page f Profiles API — Use the HCL Connections Profiles API directly to read and write extended profile attributes programmatically. The three approaches complement each other — automated for HR data, user-driven for personal expertise, and API-driven for programmatic control. 

Customization workflow Define custom fields in profiles-config.xml 

Customization workflow Define Custom Control Introduce custom CSS classes Change Base layout 

Customization workflow Create custom String Bundles 

Customization workflow Define where they are shown 

Customization workflow Create an Appreg application 

03 — AI INTEGRATION AI-Powered Integration 

The AI Conversation Everyone Is Having Cloud AI tools are genuinely powerful — but enterprise organizations face a consistent barrier. Data Privacy Where does your internal content actually go when it hits a cloud LLM? Compliance GDPR, data residency requirements, and industry regulations make public SaaS LLMs a non-starter. The Question Can we have AI capabilities AND keep our data fully sovereign? 

The Answer: Sovereign, On-Premises AI Open-Weight Models Are Here Llama, Mistral, DeepSeek, Gemma, GPT- OSS, Qwen, Kimi, GLM — serious, capable models with publicly available weights you can run on your own hardware. Runs entirely within your own infrastructure No data leaves your network — not a single token Performance comparable to commercial models for enterprise tasks like summarization and Q&A Cutting-edge AI Capabilities with Privacy by Design 

Open-weight vs. Open-source models 

Native Connections AI Integration Embedded directly in the Connections UI — no separate tool, no context switching. Document & File Analysis Upload a file and ask the assistant to find specific information through semantic search across its content. Intelligent Search Ask the assistant to search Connections for you and generate detailed reports — directly from the chat interface. More on Our Roadmap Additional AI-powered capabilities are in active development — stay tuned. 

Live Demo — AI Integration 04:59 Vimeo SonnyDemoVid Entirely on-premises. Entirely private. No data leaves the network. 

AI Integration Demo Summary 1 Analyzed confidential lab notes in .docx, directly out of HCL Connections 2 Created a market entry proposal from insights 3 Searched HCL Connections for internal roles (confidential) 4 Created a job description for LinkedIn for a missing role 

Architecture — Full On-Premises Stack UI / Frontend User-Interface with live token streaming API & AI Middleware Database, Cache, Backend with REST / Websocket Endpoints Local LLM Runtime Local LLM runtime (Ollama / vLLM) on GPU server Modular by Design HCL Connections integration via REST / WebSocket AI Middleware routes requests to the local LLM runtime (Ollama) Runtime manages multiple open-weight models on your GPU server Entire stack can sit within your network perimeter, how you want it. No egress. No cloud dependency. No vendor lock-in. 

Bring your own Hardware / LLM Inference Bring your own GPUs, or Ollama instance, or DominoIQ, it is modular. 

"You don't have to choose between AI capabilities and data sovereignty anymore." 100% Data Stays On-Prem Zero data exposure — not a single token leaves your network. 100% In your Workflow Embedded natively in Connections — no tool switching required. 100% Open-Weight Models Models genuinely competitive with commercial offerings for enterprise tasks.* *not restricted to open-weight, but everything you saw was powered by open-weight models. 

04 — BLOGS API Modern Blogs API 

APIs Built for the Modern Developer HCL is rebuilding the Connections API layer from the ground up. The Blogs API is the first live service via the new API Gateway — preview available now. Old Blogs API New Blogs API Format XML / AtomPub Clean JSON Auth Basic / Cookie OAuth 2.0 Documentation Static wikis Live OpenAPI / Swagger Developer UX Complex and painful Clean and straightforward More services to follow — Communities, Profiles, Files and beyond. 

What You Can Build With the New API Content Widgets Pull blog posts into any modern intranet page with a simple fetch call. Mobile Apps Read and write Connections blog content natively in iOS or Android applications. Content Syndication Push blog posts from Microsoft Teams, SharePoint, or external sites automatically. AI Integration Use on premise AI to generate your blog posts and publish them directly 

Live Demo — Swagger UI The OpenAPI spec for the Blogs API documents every endpoint — list blogs, fetch entries, create content — and lets you test calls directly from the browser. This is what modern developer experience looks like. No XML parsing. No ancient wikis. Just clean, self-documenting APIs. 

Key Takeaways Connections Is More Extensible Than You Think Modern tooling works with Connections — not against it. You don't have to wait for HCL to ship features. Your Data Is Your Biggest Asset — Surface It Better Org structures, HR attributes, and content are already there. They just need a better interface. Sovereign AI Is Real and Accessible Don't let data privacy concerns stop AI exploration. On-prem models are capable, practical, and deployable today. HCL Is Investing in the Platform The new API layer signals a strong future for Connections. Now is a great time to build. 

Questions? 

Open-source vs. Proprietary models