Softwarefor health, growth, andwhat's next.

UltraFit360 is a cross-platform React Native application written in TypeScript, shipped to both Apple App Store and Google Play from a single codebase. The app uses Firebase Auth for sign-in, Firestore as the live database for workouts, programmes and progress, and Cloud Functions for any privileged write. Realtime sync between devices is handled by Firestore's native listeners, so progress saved on the watch shows up in the app instantly. Audio coaching uses ElevenLabs streaming voices generated on-demand from a coach's text script. Crash and performance monitoring run through Firebase Crashlytics + Performance. UltraFit360 launched in 2026 and is currently live with 5,000+ Android installs, 700+ iOS installs and a 4.8★ App Store rating after six months in market.

The UltraFit360 Coach portal is a separate React + TypeScript web application that gives trainers a programme-builder, member messaging and analytics view alongside the consumer app. It shares the same Firestore database as the mobile app, with security rules that scope each coach to their own roster — coaches can't read another coach's members. The portal is deployed on Firebase Hosting behind Firebase Auth, with role-based access enforced server-side via Cloud Functions. Programme drafts use Firestore transactions so two coaches editing the same programme don't overwrite each other. Charts are rendered with Recharts. The portal is in active development; the alpha is being trialled with the same coaches whose members already use UltraFit360, so the two products evolve together rather than diverging.

The studio runs an internal pipeline that turns a one-line brief into a finished Instagram Reel, a multi-slide carousel and a caption — all without manual editing. A scheduled Cloud Run worker calls a large language model to draft a script, breaks it into shot beats, and renders each shot from a stock-clip library or generated b-roll. Voiceover is synthesised by ElevenLabs with the studio's cloned brand voice and timed against the script. Slides for carousels are composed with a headless browser rendering a templated HTML page to PNG so typography stays on brand. Output drops into Cloud Storage and posts via the official Instagram Graph API. End-to-end runtime is roughly four minutes per asset. The same pipeline powers AptixLabs' own social presence and is the basis for the studio's social-automation offering for partners.

AptixLabs uses ElevenLabs as its voice partner across every product that talks. For UltraFit360 the studio uses ElevenLabs streaming voices to deliver real-time audio coaching during workouts — text scripts written by coaches are turned into spoken audio on-device-trigger, so the same coach voice can cover thousands of new workouts without re-recording. The studio also clones a brand voice for the automated reel pipeline so all generated content sounds the same, regardless of who wrote the script. Latency-sensitive paths use ElevenLabs' streaming endpoints; longer-form output is generated ahead of time and cached in Cloud Storage. ElevenLabs partnered with AptixLabs early and contributed 33 million character credits, which is enough to ship lifelike audio coaching across the full UltraFit360 rollout without billing pressure.

AI agents at AptixLabs are application-specific, not generic chatbots. The studio builds them around a clear job-to-be-done — onboarding a new gym member, answering questions about a programme, generating a workout from a template — and grounds each agent in the client's own Firestore data via tool calls. The agent runtime is built on the Anthropic and OpenAI APIs depending on the strengths needed; for cost-sensitive paths the studio routes to Gemini via Vertex AI. Agents run inside Cloud Run and stream responses to the client. State is persisted in Firestore so a user can return to the same conversation across devices. Every tool call is logged and a guardrails layer enforces what the agent can and cannot do, so an agent helping a fitness member can never accidentally write to billing or admin data.

UltraFit360 reads from Apple HealthKit on iOS and Google Health Connect on Android using their native SDKs through React Native bridge modules. Supported signals include workouts, heart rate, sleep and active energy. The data is fetched on app open and on a background schedule when the OS permits, then written to a per-user Firestore document so the coach portal can see it without re-syncing. Apple Watch integration uses the WatchConnectivity framework for live workout sessions; Garmin and WHOOP are supported through their public APIs as opt-in connections. Wearable data is treated as health data — encrypted at rest, scoped to the member who owns it and never used for model training. The studio considers wearable integration table stakes for any fitness app shipping in 2026, because users abandon apps that can't see what their watch sees.

AptixLabs builds health applications under a compliance-first architecture. UltraFit360 minimises personally identifiable health information to the smallest viable schema, encrypts data at rest in Firestore with Google-managed keys, and routes every write through Cloud Functions so the audit trail is centralised. Business Associate Agreements are signed with subprocessors that touch protected data. The privacy policy maps every data category to a lawful basis under GDPR and a corresponding clause under India's DPDP Act 2023. Access reviews run quarterly. For clients who need a stronger posture, the studio can deploy into a customer-owned Google Cloud project with VPC Service Controls and customer-managed encryption keys. The studio is not currently SOC 2 certified, but the controls align with the SOC 2 trust principles and the path to certification is documented for any partner who needs it.

The studio defaults to Firestore in Native Mode as the primary database. Firestore is a serverless document store with real-time listeners, fine-grained security rules and global replication out of the box — it removes most of the operational work that comes with running a Postgres cluster. For analytics, query-heavy reporting or any workload that needs SQL joins, the studio exports to BigQuery via the Firestore-to-BigQuery extension and runs analytical queries there. Hot paths are denormalised; rarely-touched fields stay normalised. The combination handles UltraFit360's current traffic at a low double-digit dollar cost per month and scales to the millions of documents tier without any architectural rewrite. The studio reaches for Cloud SQL Postgres only when a project genuinely needs SQL constraints from day one — most products don't.

Real-time sync uses Firestore's native onSnapshot listeners on every client surface — the React Native app, the React coach portal, and any admin tools the studio builds. When a member finishes a set on the watch and it writes to Firestore via the phone, the same record streams to the coach's open dashboard within roughly 100–300 milliseconds, depending on network. There's no separate WebSocket server, no Redis pub-sub, no Pusher channel — Firestore is the message bus. Security rules ensure each listener only receives data scoped to that user's tenant. For features that need stronger ordering guarantees, like programme edits with conflict resolution, the studio uses Firestore transactions. The model lets a three-person team operate a real-time product without an infra engineer on payroll, which is a deliberate choice for the studio's stage.