Context-Aware Business Search
Hybrid lexical + dense retrieval with cross-encoder re-ranking for organizational knowledge. Measuring gains over keyword search on internal corpora.
ExploreWe turn uncertain technology into working systems.
DELRIQUE Labs is an applied research lab. We explore emerging tech, run disciplined experiments, build the ideas that survive — and customize open-source systems so organizations can own what they run.
01 — The Lab
A place to test what technology can become.
Technology moves fast. The lab gives us room to investigate that change through practical work — building prototypes, testing new tools, and documenting what we learn, good or bad.
Explore
Understand a technology, a problem, or an emerging idea before committing to it.
Experiment
Test it in a controlled environment with clear hypotheses and measurable outcomes.
Build
Turn the ideas that survive into working prototypes and reusable tools.
Share
Document, publish, or open-source what can benefit others.
Learn
Feed every outcome — success or failure — back into the next cycle.
Govern
Apply privacy, security and control as design constraints, not afterthoughts.
02 — Focus Areas
What we are actively investigating.
These are questions, not services. Each area has running projects behind it.
AI & Intelligent Systems
Local AI, retrieval pipelines, automation, and AI-assisted workflows with human oversight.
6 projectsSoftware & Engineering
Developer tools, interfaces, and practical digital utilities built and battle-tested.
4 projectsAutomation
How repetitive digital work can be simplified with software and intelligent systems.
activeOpen Source
Reworking, extending and publishing useful technology for others to build on.
3 reposPrivacy & Control
Technologies that give people transparency and ownership over their digital systems.
principleEmerging Technology
Investigating new platforms, models and frameworks before they go mainstream.
scanning
03 — Projects
What we are building.
Projects are ideas that have earned their way past the experiment stage. We build them deliberately, test assumptions as we go, and publish what proves useful.
Context-Aware Business Search
01 / Tested
Can organizational search understand business context?
Keyword search matches strings, not meaning. We asked whether results improve when a system weighs entities, relationships and intent.
02 / Hypothesis
Semantic ranking beats lexical matching
Results improve when structured organizational data is combined with semantic ranking and evaluated against internal test corpora.
03 / Approach
Hybrid retrieval + re-ranking
We prototype pipelines that fuse lexical and dense retrieval, then re-rank with a lightweight cross-encoder tuned on domain documents.
04 / Result
Findings maturing
Early runs show measurable gains on internal corpora. Full write-up publishes as the experiment matures.
On-Device Inference Pipelines
Running small language models close to the user and infrastructure — latency, cost and privacy trade-offs for enterprise use.
ExploreResilient Workflow Automation
Idempotent, observable automation for repetitive digital work — with human checkpoints where judgement matters.
ExploreStructured Organizational Memory
Capturing institutional knowledge as structured, queryable graphs rather than documents that decay.
ExplorePrivacy-First AI Workflows
Frameworks that keep data control, transparency and human oversight fundamental in AI-assisted processes.
ExploreSelf-Hosted Enterprise Tooling
Modular, interoperable components that return control of the stack to the organization.
Explore
04 — Open Source
Built to be shared.
What we learn doesn't stay locked up. We publish reusable libraries, tools, and infrastructure for others to build on.
Research Libraries
Reusable libraries and notebooks that package what we learn about AI, retrieval, and evaluation for others to build on.
Developer Tools
Small, focused utilities that remove friction from daily engineering work — clis, helpers, and integrations.
Infrastructure
Composable infrastructure components for self-hosted, sovereign systems that scale with your organization.
05 — Services
Open source, customized for your organization.
Don't build from scratch. Don't pay proprietary lock-in. We take proven open-source projects, harden them, theme them to your brand, and integrate them into your stack.
Hardened by default
Security, privacy, and reliability treated as engineering constraints from the first commit — not bolted on at handover.
Themed to your brand
Every interface reworked to match your identity so open source feels like it was built for you, not repurposed.
Integrated, not isolated
Wired into your auth, data, and workflows rather than left as a disconnected tool your teams must learn around.
01
Fork
Start from a proven open-source base rather than a blank page.
02
Harden
Apply security, privacy, and reliability hardening.
03
Theme
Customize the interface to your brand and workflows.
04
Integrate
Connect to your existing auth, data, and tooling.
05
Extend
Add the features your organization truly needs.
06
Document
Produce runbooks and maintainer documentation.
07
Maintain
Keep it patched, updated, and running smoothly.
08
Hand over
Give your team full ownership and source control.
Slow & costly
Build from scratch
- Months of effort
- Ongoing maintenance burden
- Security debt from day one
- Expensive to iterate
Locked in
Proprietary SaaS
- Recurring license costs
- Vendor lock-in
- Limited customization
- Data leaves your control
Recommended
Customized open source
- Proven, battle-tested base
- Full source ownership
- Themed to your brand
- Your data stays yours
Enterprise tooling
Project management, collaboration, and internal platforms your teams will actually adopt.
Internal utilities
Templates, dashboards, and helpers that remove daily friction across departments.
Observability & dashboards
Monitoring, logging, and metrics that give you a clear view of your systems.
Automation layers
Repeatable pipelines and workflows that reduce manual, error-prone work.
1
Audit
Understand your stack, constraints, and what you need to own.
2
Plan
Define scope, architecture, and a realistic timeline with milestones.
3
Customize
Harden, theme, and integrate the chosen open-source base.
4
Integrate
Wire into your auth, data, and workflows with runbooks.
5
Handover & Maintain
Document, train, and support with clear ownership for your team.
06 — Insights
From the lab.
Notes and findings from our research as it happens.
AI Observability Meets the Enterprise: Production Oversight for 2026
Running AI in production means shift from valuing what launches to governing what runs. Observability gives teams the control to detect, debug, and improve models at scale.
The Shift Toward Self-Hosted Privacy-First Enterprise Workspaces
Why corporate enterprises are revisiting data sovereignty and adopting local AI models for internal knowledge management.
RAG Is the Enterprise AI Pattern of 2026: Retrieval-Augmented Knowledge Engines
Retrieval-augmented generation grounds AI answers in an organization’s own governed knowledge — cutting hallucination risk while keeping data inside the perimeter.
Agentic AI: When Machines Complete Entire Workflows, Not Just Responses
Agentic systems plan, execute, and verify multi-step workflows. With them come new identity, audit, and governance questions for non-human workers.
Have something worth exploring?
Bring us a hard problem. We'll investigate it with you — and build what comes out of it.