Portfolio builderRust-first backend • Operator-built • Long-term focus
We build software businesses.Built for long-term operation.
Kyndrid is an umbrella brand for a portfolio of products — some built internally, some built with partners. We build for long-term operation: products with clear economics, measurable execution, and systems that don’t require heroics, with Rust-first backend development.
Backend
Rust-first
Reliability
Retries, idempotency, clear failure modes
Compliance
Audit trails + retention-aware design
Ops
Metrics, alerts, and runbooks
MediaWorkflow automationComplianceERP / MRPApplied AIInfrastructure
Portfolio highlights
Built to own and operate
Kyndrid (founded 2017) builds products for long-term operation — not contract development.
Sellarity Studio is shipping soon
Best-in-class studio management: scheduling, memberships, billing, POS, and client lifecycle automation — built for real operators.
Engineering discipline that survives scale
Measurable ops, security boundaries, and failure-aware design so systems stay stable under stress.
Portfolio includes long-running operators
E-File Magic has been in operation since 2006, supporting high-volume filing workflows for thousands of customers and the businesses they serve.
Built in production
These platforms live in real environments: audits, deadlines, integrations, and the parts of a business that can’t break. We build for operation and longevity.
RustReactAngularAWSServerless
Portfolio
Companies and platforms
A snapshot of products we build and operate — some built internally, some built with partners. Details vary by scope and confidentiality.
University of Fighting
Sellarity ERP
Sellarity Studio
E-File Magic
University of Fighting
Online training platform for MMA/BJJ and combatives — built for long-running content libraries and real operations.
Venture
University of Fighting, LLC
Stack
Backend: RustFrontend: React / AngularAutomation: Assistive toolingPlatform: Sellarity base
Highlights
- Media portal built for real content pipelines (courses, clips, curriculum progression).
- Assistive tagging, search, and content structuring to reduce ongoing overhead.
- Architecture designed for reliability under spikes and long-term growth.
MediaOperationsScalePartner venture
Sellarity ERP
A full-scope ERP with accounting at the core — built for auditability, multi-entity operations, and long life.
Venture
Kyndrid
Stack
Backend: RustFrontend: React / AngularAutomation: Integrated toolingDomains: Accounting, Sales, Purchasing, Inventory, Manufacturing, E-Commerce
Highlights
- Sales + fulfillment: quoting → ordering → pick/pack/ship, retail sales, returns, and refunds.
- E-Commerce: catalog, pricing rules, taxes, discounts, and order ingestion/fulfillment sync.
- Purchasing + procurement: vendors, POs, receiving, landed-cost patterns (where used), and approvals.
- Light manufacturing: BOMs, work orders, component consumption, and inventory movement with traceability.
- Multi-currency + multi-location operations with transfers and inventory valuation integrity.
- Full accounting suite: A/R, A/P, General Ledger, Financial Statements, General Journal, journals, and intercompany flows.
Core platformFull accountingSales + fulfillmentPurchasingE-CommerceManufacturingMulti-currencyTransfers
Sellarity Studio
Best-in-class studio management for serious operators — scheduling, memberships, billing, POS, and client lifecycle automation in one system.
Venture
Kyndrid
Stack
Backend: RustFrontend: ReactInfrastructure: AWS LambdaDomains: Scheduling, Memberships, Billing, POS, CRM
Highlights
- Scheduling for classes + appointments: resources/rooms, instructor assignment, waitlists, check-in, attendance, and no-show controls.
- Memberships + packages: autopay, proration, invoices/receipts, discounts, promo codes, and payment retry/dunning.
- Point-of-sale + retail: product catalog, taxes/discounts, and on-site checkout workflows (with inventory awareness).
- Client CRM: profiles, tags, notes, segmentation, and lead → member pipeline tracking with measurable follow-up.
- Comms + reminders: email/SMS touchpoints, confirmations, reminders, and reactivation sequences.
- Built for reliability: audit trails, idempotent sync + retries, and clean failure modes around third-party integrations.
Best-in-classSchedulingMembershipsBillingPOSCRMAutomation
Established platform
E-File Magic
A long-running, high-volume compliance platform supporting large-scale 1099 operations. Built and operated through real filing seasons, with emphasis on correctness, auditability, and operational visibility.
- Compliance-heavy workflows and evidence trails driven by real-world deadlines.
- High-volume data processing pipelines with strong validation and error isolation.
- Operational tooling and observability designed for peak windows and low error budgets.
ComplianceHigh volumeAudit trailsOperational toolingOperating since 2006
Domains
Where we build
The problem spaces we work in across the portfolio — the kind that need reliability, auditability, and clean data flow.
Financial Technology
Transactional workflows, reporting pipelines, reconciliation, and correctness-first processing — designed for peak windows.
ERP / MRP
Multi-company domain modeling, inventory/manufacturing flows, purchasing/sales, and ledger integrity — built for long life.
Accounting + Compliance
Audit trails, evidence artifacts, retention-aware storage patterns, controlled access, and operational visibility for regulated workflows.
Customer Intake + Lifecycle Automation
Lead ingestion, onboarding, routing, reminders, and integration reliability — measurable outcomes, not guesswork.
Applied AI (practical)
Search, extraction, routing, and assistive tooling — implemented with validation, fallbacks, and monitoring.
Infrastructure Engineering
Serverless, containers, edge delivery, resilience design, and observability — built to keep outages short and rare.
How we build
A straightforward approach
Define what must be true, design for operation, ship measurable systems.
Step 1
Define invariants
Correctness, security boundaries, and what must always be true.
Step 2
Design for ops
Retries, idempotency, visibility, and sane failure modes.
Step 3
Build in Rust
Rust on the backend, with strict boundaries and predictable behavior.
Step 4
Ship measurable
Metrics, logs, traces, and evidence trails you can use.
Applied AI
Automation with constraints
We use AI to reduce manual work — search, extraction, routing, and assistance — while keeping accuracy, privacy, and auditability under control.
Speech-to-text pipelines
Audio → transcription → formatting, with timestamps that fit into real workflows.
- Streaming + batch modes with backpressure and queueing.
- Vocabulary and formatting rules where supported.
- Quality monitoring with graceful fallback behavior.
Low latencyTimestampsBatch + streaming
Similarity search
Find the right document, customer, or record quickly — tied to sources and tuned to the corpus.
- Index strategies matched to query patterns and data shape.
- Filters and reranking to reduce junk matches.
- Caching and spend controls for predictable costs.
Source-linkedQuality controlsFast retrieval
Private assistants
Assistants with predictable behavior — constrained by schemas, access controls, and logged decisions.
- Structured outputs + strict validation on every response.
- Least-privilege access to data and actions.
- “I don’t know” is acceptable (and useful) when logged.
Structured outputsAccess controlsPredictable behavior
Document analyzers
Forms, PDFs, and messy inputs — extracted with schemas, confidence gating, and review lanes when needed.
- Schema-driven extraction with confidence thresholds.
- Human review paths where stakes exist.
- Retention-aware storage and audit trails.
OCR / formsConfidence gatingAudit trails
How we ship AI safely
- Structured outputs (schemas) + strict validation on every response
- Least-privilege access with allow-listed actions and retention-aware storage
- Defenses against malicious inputs and data leakage in retrieval workflows
- Human review gates where stakes exist (money, compliance, identity)
- Fallbacks and circuit breakers when confidence drops or systems degrade
- Replayable inputs + audit logs for debugging and governance
Eval harnessCanariesRollback switchEvidence trails
The contract (boring on purpose)
input → normalize → retrieve → infer → validate → fallback → persist → monitor
Models are components. We constrain them with validation, access controls, monitoring, and fallbacks — so the system behaves predictably even when the model doesn’t.
Latency budget
Concurrency, caching, backpressure
Quality budget
Evals, drift, confidence gating
Safety budget
Access controls, review lanes, logging
Cost budget
Caps, batching, model selection
Contact
Partnerships + product
For product demos, integrations, and partnership ideas, reach out. We don’t take services engagements; we collaborate when there’s long-term alignment.