The software SDEN designed, built, and
runs in production.
SDEN designs, builds, secures, and operates four SaaS products: Real Estate (real estate), Lead Manager (sales automation), Beauty (salon management), and University Portal (education). They are the engineering proof of what we deliver to clients.
Four SaaS products
SDEN's own SaaS products are the proof of the engineering we sell. Each one is designed, built, secured, and operated by SDEN end-to-end — and used in production by real businesses. We do not ship marketing decks as evidence; we ship products you can navigate to, use, and form your own opinion about.
The four products below cover four different sectors (real estate, sales automation, salon and wellness, higher education) and four different operational shapes (multi-tenant B2B SaaS, AI-augmented workflow tool, schedule-driven point-of-sale, role-based academic platform). The technical surface is wide on purpose. The lesson is the same: the production discipline is the engineering — not the framework, not the stack, not the slide deck.
Why we
run our own products.
Consulting can teach you to recommend. Only operating production software teaches you what survives. The reason SDEN runs its own SaaS products — even though we are a services-led company — is to keep our engineering honest against the pressures that only show up after launch: real customers paying real money, support tickets that arrive faster than the engineering team can plan around, security disclosures from researchers who do not care about our roadmap, and the slow accumulation of small architectural decisions that decide whether the product can still ship a feature three years from now.
Every client engagement inherits that operational scar tissue. When we recommend a database engine, we are recommending one we have run under load. When we choose a deployment topology, it is the version we wish someone had recommended to us before we shipped the first one. When we propose a security control, the threat model behind it has been validated against actual disclosures, not just textbooks. Running our own products is the most expensive way to learn that — and it is also the most direct way to be sure we are not selling our clients a stack we would not run ourselves.
Real Estate — all-in-one real-estate platform
Real Estate unifies the whole sales cycle in one white-label product: SEO listings and an interactive map, a CRM that matches buyers to properties, AI home staging that valorizes a listing from a single photo, AI-assisted valuation, and an isolated client portal for e-signature and document tracking.
Real-estate agencies historically juggle four to six tools — a CRM, a listings spreadsheet, an e-signature service, a separate document store, sometimes a valuation tool, sometimes a property-tour booking calendar — and the work of reconciling them lives in the agent's head. Real Estate unifies those tools into a single product where listings, properties, contacts, and transactions share one data model. The agent stops re-entering the same information across systems; the agency stops losing transactions to administrative friction.
The architecture is a Next.js application backed by a typed NestJS API and PostgreSQL with per-tenant isolation. AI-assisted valuation is a Python pipeline trained on property characteristics and market history; it produces an explainable range rather than a single point estimate, and the agent's adjustment is logged so the model improves over time. AI home staging runs as a generative image pipeline — furnishing, emptying, or decluttering a room from a single photo — queued and rendered asynchronously, with a before/after comparator on the listing. The client portal is a separately hosted surface with no access to any other listing's data — multi-tenant isolation is enforced at the type level, not just at the query level.
The result is that an agency that adopts Real Estate replaces a stack of disconnected tools with one product, and the work of reconciliation disappears. The exact figures below are illustrative until we publish audited customer outcomes.
−70% PLACEHOLDER
less time per valuation
−12 days PLACEHOLDER
off the average transaction
99.98% PLACEHOLDER
measured uptime
Lead Manager — sales automation engine
Lead Manager captures leads from multiple sources, scores them with a classifier that ranks real intent over noise and explains every lead, runs multilingual follow-up sequences that stop as soon as a prospect replies, and gives each rep an AI copilot for outreach and meeting prep.
Sales teams spend more time sorting prospects than selling them. Lead Manager inverts that ratio. Leads land from forms, APIs, and webhooks into a single deduplicated store, are enriched and scored automatically against the team's ideal customer profile, and surface in a ranked queue that the rep works in priority order. Follow-up is triggered by behavior (a lead opened an email, visited a pricing page, replied on LinkedIn) rather than by a fixed cadence — and every sequence stops the moment the prospect responds, so reps never over-contact a hot lead or under-contact a warm one.
The engineering challenge is the orchestration. Multichannel sequences (email, LinkedIn, SMS, WhatsApp) have to compose without stepping on each other, the scoring model has to be evaluated continuously against actual conversion outcomes, and the system has to handle the operational realities of CRM integration — duplicate prospects, mismatched fields, partial data — without dropping leads on the floor. The stack pairs a Node.js orchestration layer with an n8n-style workflow engine, OpenAI-class classification, and a PostgreSQL store with strict tenant isolation.
The outcome teams report is fewer follow-ups, higher reply rates, and dashboards that finally tie conversion back to the source. The figures are illustrative until we publish audited customer outcomes.
×2.4 PLACEHOLDER
more meetings per hour
60% PLACEHOLDER
of follow-ups automated
99.98% PLACEHOLDER
measured uptime
Beauty brings online booking, multi-practitioner scheduling, loyalty, payments, and automated reminders into one interface designed to be run by the whole salon team — with no training.
Salons and wellness studios live and die by the calendar. Beauty treats the calendar as the source of truth: practitioners, services, resources, and slots are modeled together, double-booking is structurally impossible, and online booking shows real availability in real time rather than a stale snapshot that has to be reconciled by a phone call. Clients book when it occurs to them — outside opening hours, on a phone, in twenty seconds — and the team keeps the phone for treatments instead of for booking management.
The operational gain is in the reminder engine. Multichannel reminders (WhatsApp, SMS, email) go out on the channel each client actually reads; one-tap confirmation handles the response; an automatic waitlist re-offers a freed-up slot the moment a cancellation happens. Loyalty is built into the client record, not a separate paper card that gets lost. Payments flow through Stripe, with deposits, service history, and reconciliation attached to the client.
The result is fewer missed appointments, more bookings outside opening hours, and a salon that runs the technology rather than the other way around. The figures are illustrative until we publish audited customer outcomes.
−34% PLACEHOLDER
missed appointments
+48% PLACEHOLDER
more out-of-hours bookings
99.98% PLACEHOLDER
measured uptime
University Portal — higher-education platform
University Portal unifies the entire academic lifecycle — courses, exams, theses, internships, events, messaging, and notifications — in one role-aware platform, with a built-in AI assistant that adapts to each of five user roles.
Higher-education institutions still run academics on paper and disconnected spreadsheets: enrollment in one place, grades in another, schedules in a third, and no shared channel between administration, faculty, and students. University Portal replaces that with a single platform where courses, exams, theses, internships, and events share one data model — and where each of five roles (administrator, director, professor, student, staff) gets an interface scoped to exactly what it may see and do.
The architecture is a Next.js application backed by a typed NestJS API and PostgreSQL with Prisma. Role-based access is enforced at the type level rather than hidden in the UI, audit logs and two-factor authentication are wired in, and a real-time layer (server-sent notifications, web push, instant messaging) keeps everyone on one channel. The AI tier is a role-scoped assistant built on a hosted LLM with a curated toolset — natural-language search, document summaries, revision plans, proactive alerts — each tool restricted to the data the calling role is allowed to touch.
The result is that an institution adopting University Portal collapses hours of repetitive admin, makes a student's path traceable in real time, and gives every role an assistant grounded in real data rather than a generic chatbot. The figures are illustrative until we publish audited customer outcomes.
15
AI tools built into the assistant
5
role-tailored experiences
99.98% PLACEHOLDER
measured uptime
The production discipline
every SDEN product inherits.
The four products serve different sectors but share the same production discipline — and that discipline is what every SDEN client inherits when they hire us to build their software.
- Regional hosting by default — deploy in your region (US, Canada, or EU) on AWS, GCP, or Azure — data-residency clarity for North American customers from day one.
- Per-tenant data isolation enforced at the type level, not bolted on with hopeful WHERE clauses.
- Encrypted backups (AES-256) with restore-tested recovery procedures and an off-region copy.
- Observability wired in before launch — structured logs, RED metrics, distributed tracing, SLOs that the on-call engineer can read at 3 a.m.
- Incident response runbook with a documented disclosure path for security researchers (security@sdenengineering.com).
- Uptime target of 99.98% PLACEHOLDER measured over a rolling twelve-month window, published transparently.
What we are
building next.
SDEN is selectively expanding the product roster into sectors where we see a recurring engineering need that off-the-shelf SaaS does not serve well. Candidates under active development include vertical-specific applications in regulated industries and operational tooling for small teams that have outgrown spreadsheets but cannot justify a salesforce-scale platform. Specifics are deliberately withheld until the products are ready to operate in production — the proof philosophy goes both ways.
Work —
questions about the products.
Direct answers to the questions we get asked the most. If yours isn't covered, write to the team.
Paste your URL. Get an
AI audit in 20 seconds.
Most agencies talk about AI audits. Run one right now — we'll check performance, SEO, AI-readiness, and security headers, then narrate the result.
Streaming findings will appear here.
Paste a URL on the left and hit Audit my site.
Got a project worth building?
Tell us about your project. We work with a limited number of clients at a time — and we'll get back to you within 24 working hours with a first engineer's read, no commitment.