LoadingSoftware that carries weight.ONYXERA TECHLoading route
Web platforms, mobile apps and AI systems for teams that cannot afford downtime. We ship them, instrument them, and stay on the pager afterwards — same engineers, no handoff.
Screenshot of the Onyxera delivery console: p95 latency 142 ms and falling, error rate 0.03 percent, 34 production deploys per week, all health checks passing.
No partner network, no subcontractors, no handoff to a team you have never met. The people who scope your architecture write the migrations, ship the app, and answer the pager at 3am.
Most engagements start at 01 or 05 and end on 06. Nothing is subcontracted.
One typed contract, from the database to the button.
Server components where they earn their keep, client components where they don't — decided per route, not per fashion cycle. The design system ships as a versioned package with its own changelog, so a token change is a pull request instead of a Slack thread. Commerce runs headless because merchandising should never be able to block a deploy.
export async function GET(req: Request) { const q = OrderQuery.parse(searchParams(req)) const rows = await db.query.orders.findMany({ where: and(eq(orders.status, q.status), lt(orders.id, q.cursor)), orderBy: [desc(orders.placedAt)], limit: 50, }) return Response.json({ rows, next: rows.at(-1)?.id ?? null })}Stack Next.js · React · TypeScript · Tailwind CSS · tRPC · Drizzle · PostgreSQL · Vercel
Result LCP 4.1s to 1.2s, checkout completion up 34%, inside one quarter and without a redesign.
Full specificationOne typed contract, from the database to the button.
Server components where they earn their keep, client components where they don't — decided per route, not per fashion cycle. The design system ships as a versioned package with its own changelog, so a token change is a pull request instead of a Slack thread. Commerce runs headless because merchandising should never be able to block a deploy.
export async function GET(req: Request) { const q = OrderQuery.parse(searchParams(req)) const rows = await db.query.orders.findMany({ where: and(eq(orders.status, q.status), lt(orders.id, q.cursor)), orderBy: [desc(orders.placedAt)], limit: 50, }) return Response.json({ rows, next: rows.at(-1)?.id ?? null })}Stack Next.js · React · TypeScript · Tailwind CSS · tRPC · Drizzle · PostgreSQL · Vercel
Result LCP 4.1s to 1.2s, checkout completion up 34%, inside one quarter and without a redesign.
Full specificationShared where it pays. Native where it has to be.
React Native on the new architecture handles the ninety percent that is lists, forms and navigation. Swift and Kotlin handle the camera, the Bluetooth stack and anything that touches the background scheduler, because bridging those is a false economy. The app assumes the network is lying to it, so it still works in a lift, a basement, or a warehouse on two bars.
Stack React Native · Expo · Swift · Kotlin · TypeScript · WatermelonDB · Fastlane · Sentry
Result 1.9s cold start on a five-year-old Android mid-ranger. 99.4% crash-free sessions.
Full specificationEvals first. A good demo is not evidence.
Retrieval is a search problem with a language model bolted on the end, and we build it in that order. Every prompt, model and chunking change runs against a 600-case golden set before it merges, and the scores are in the pull request where an engineer has to look at them. Cost per session and p95 latency are acceptance criteria, not something you discover on the invoice.
const suite = defineSuite({ dataset: golden("support-qa", { cases: 600, seed: 7 }), run: ({ question, tenantId }) => pipeline.answer(question, { tenantId, k: 24, rerankTo: 4 }), scorers: [factuality, citationRecall, refusal, costPerCase],}) // CI blocks the merge if any threshold regresses against main.export default suite.gate({ factuality: gte(0.92), citationRecall: gte(0.9), p95LatencyMs: lte(1800),})Stack Python · TypeScript · LangGraph · pgvector · Qdrant · Braintrust · Modal · Temporal
Result Accuracy 71% to 94% across 600 graded cases, at $0.04 median cost per session.
Full specificationIf you can't rebuild it from an empty account, you don't have it.
Everything is Terraform. No console clicks, no environment that only one person understands, no resource whose origin is a mystery to the team that owns it. Deploys go out on merge behind a 5% canary, and the rollback is one command that has been rehearsed on a Tuesday. We instrument before launch so the dashboard tells you what broke before a customer has to.
Stack AWS · Google Cloud · Terraform · Kubernetes · Docker · GitHub Actions · OpenTelemetry · Grafana
Result Deploy lead time 6 days to 40 minutes, with monthly spend down 38% and no capacity lost.
Full specificationWe don't do discovery theatre.
Fourteen days turns a pile of opinions into a scope with numbers on it, including the parts we think you should cut. Prototypes go in front of eight real users before production code exists, because a stakeholder preference is not a finding. The system ships as tokens and coded components — what is in Figma is what is in the build, or the build wins and Figma gets corrected.
export default defineTokens({ source: "figma/variables/*.json", transforms: ["name/kebab", "color/oklch", "size/rem"], targets: [ { platform: "css", out: "packages/ui/tokens.css" }, { platform: "swift", out: "ios/Sources/Tokens.swift" }, { platform: "compose", out: "android/ui/Tokens.kt" }, ], // A token no component references fails the build. lint: { orphans: "error", contrast: "AA" },})Stack Figma · Figma Variables · Storybook · Style Dictionary · Maze · Axe DevTools · Rive
Result Onboarding drop-off down 46% and support tickets down 31% in the first month after launch.
Full specificationWe stay on the pager after launch.
Every product we ship can move onto an agreement with a named lead engineer, a fifteen-minute P1 response and a monthly review that has real numbers in it. Dependencies get patched on a schedule instead of when a scanner starts shouting. Retainer hours go into a shared backlog, so the product keeps compounding rather than quietly rotting for two years.
resource "datadog_monitor" "checkout_p95" { name = "checkout p95 over 400ms (5m)" type = "metric alert" query = "p95:trace.http.request{service:checkout} > 0.4" monitor_thresholds { warning = 0.3 critical = 0.4 } notify_no_data = true escalation_message = "P1 — page the named lead engineer"}Stack PagerDuty · Datadog · Sentry · Snyk · Dependabot · Statuspage · Linear · Terraform
Result 99.98% uptime across 40 managed services. Median P1 acknowledgement: 11 minutes.
Full specificationMost engagements open with a 14-day discovery sprint and close on a support agreement.
Sampled hourly from production telemetry. Availability excludes scheduled maintenance windows.
Five phases, fixed rituals, a written artefact at every gate. You get a scoped estimate before anyone opens an editor and a preview URL on every pull request from the first build sprint.
Interviews with the people who will actually use it, a technical audit of whatever already exists, and a scope written in ranges rather than one hopeful number. You leave with a plan specific enough to hand to another agency — which is the point.
The data model, the service boundaries and the interfaces on top of them, argued out in one room and then written down as decision records. Prototypes go in front of users while those records get signed, so the expensive disagreements happen now and not in month four.
Two-week sprints against a board you can read at any hour, each ending in a deploy rather than a status update. Every pull request runs typecheck, unit, integration and end-to-end suites, then publishes a preview URL you can click before anyone approves it. If a sprint slips we tell you on the day it slips, with the reason and the new date.
Load tests to 3x projected peak, a penetration test, accessibility remediation and a rollback we have already rehearsed against production data. Launch day is a checklist read out loud by one person, not an event the team has to survive.
An SLA-backed agreement with a named engineer, monthly patch cycles, and a reliability review with real numbers in it. Retainer hours go straight into the roadmap, so year two makes the product better instead of merely keeping it alive.
Six engagements, described by the people who ran the migration, carried the pager and reviewed the diffs. One of them talked us out of the build they came in asking for.
Written references · figures reconciled against client dashboards
“We asked for a chatbot. They talked us out of it, shipped a retrieval index in six weeks, and it now closes 71% of tickets without a human.”
Quotes edited for length, nothing else. Latency and release figures come from the client’s own telemetry, not ours.
“Two studios quoted nine months for the replatform. Onyxera cut us over in eleven weeks behind a dual-write window, with a rehearsed rollback at every step. We never took the storefront down.”
0 s planned downtime · +23% checkout
“Our field techs work in hospital basements with no signal, so offline-first was not a nice-to-have. They rewrote the sync layer twice to get conflict resolution right, property-tested it, and billed us for the first attempt only.”
0 sync data-loss reports in 14 mo
“Releases used to be a Friday-night ritual: four people, one spreadsheet, and a rollback plan nobody trusted. Ten weeks in we ship on merge, behind flags, at 1% first. I have stopped checking Slack on Saturdays.”
2 releases/mo → 40/wk · 0.4% change-fail
“The build was fine. The thirty-one months since are why I keep handing out their number — our payment provider deprecated an API on six weeks' notice and they had a migration plan and a shadow-traffic test running before we finished reading the email.”
31 months retained · 0 failed cutovers
“They carried our pager for four months and then handed it back properly: runbooks, alert thresholds we could actually defend, and a burn-down of every page they had ever answered. Read the diff, not the invoice — that is where their work shows.”
19 pages/wk → 2 · on-call handed back
Not covered here
Skip the form. We’ll put an engineer — not a salesperson — on a 30-minute call.
Talk to usTwo models. Fixed scope for well-defined projects, priced after a two-week discovery sprint, and monthly retainers for continuous product work — one squad, one flat rate. Most builds land between $60k and $400k. Discovery is $12k and credits in full against the build if you carry on. We do not bill hourly; hourly billing charges you more the slower we go, which is a strange thing to agree to.
Send the brief, the half-built repo, or the problem nobody has scoped yet. You get a written assessment — architecture, team, timeline, cost — inside five working days.