AI & Automation2025
AI RFP Proposal Platform
A retrieval-augmented proposal generator that drafts tailored RFP responses from a company's own past proposals and client requirements.
ReactFastAPIPythonMySQL
Read the issueCase files · Proof, not promises
AI case studies, with the screens to prove it
A retrieval platform that drafts proposals, an offline assistant that keeps meetings on the phone, and the CRM that makes real-estate data usable. Real constraints, real stacks, real outcomes.
The evidence
- Projects shown
- 3 shipped
- RFP turnaround
- 60% faster
- Meeting audio sent to cloud
- 0 bytes
RAGOn-device AIOpen source
Two AI builds, two hard constraints
The model was the easy part
One project had to fit inside a proposal team's existing review process. The other could not send a single byte to a server. Both constraints shaped the architecture more than the choice of model did.
AI & Mobile2026
Private AI Assistant
A free, open-source Android app that records meetings, transcribes them with on-device Whisper, summarises them with a local Llama model and answers questions about your meeting history — without sending a single byte to the cloud.
KotlinAndroidWhisperLlama 2.5
Read the issueRAG proposal platform
A React workspace over FastAPI microservices retrieval. Relevant past proposals are pulled from a MySQL document store, and the team edits a cited first draft instead of starting from a blank page.
Full case studyPrivate on-device assistant
Whisper transcribes, a local Llama model summarises and a local index answers questions — all on an Android phone. No account, no upload and no internet connection required.
Full case studyThe unglamorous foundation
AI needs connected data first
Before a model can answer anything useful, the business needs one source of truth. This real-estate suite is that groundwork: properties, leads, appointments, attendance, payroll and WhatsApp updates in one system.
We include it in this collection deliberately. The work is automation rather than a model showcase, but it is the same engineering problem an AI assistant inherits: permissions, a data model and a reliable record of what happened.
Real Estate2025
Anshwar CRM & ERP
An all-in-one CRM and ERP for real-estate teams: properties, leads, appointments, attendance, payroll and WhatsApp updates, on web and mobile.
Next.jsFlutterCodeIgniter REST APIPostgreSQL
Read the issueHow we report a case study
Numbers, not adjectives
A project only becomes a case study when there is something measurable to say. Here is what we look for.
Accuracy, measured
Every AI build gets a test set taken from real documents before launch. We report what it gets right, what it refuses and where it still needs a person.
Time saved, not vibes
The RFP platform cut proposal turnaround by 60%. That is the kind of number we ask for — measured against the old process, not a demo script.
Cost per task
Model calls, hosting and review time are part of the design. A feature that is accurate but unaffordable is not finished.
What you will not find here: invented client logos, stock-photo teams or a demo video passed off as a product. If a workflow is not live, we say so.
Read the method
The guides behind the builds
How RAG worksRetrieval, citations, evaluation and when fine-tuning is the wrong tool.Read moreWhat an AI app costsPrototype, production and the running costs that are not in the quote.Read moreWorkflows worth automatingWhere small teams recover hours first, and where a human stays in the loop.Read more
Kaboom!
Ready to start your project?
Tell us what you want to build. We will reply within one business day with honest advice and a clear next step — no pressure, no jargon.