This is what we hand you: a working, multi-tenant commerce intelligence platform for BigCommerce — plus the engineering team that built it, can extend it, can maintain it, and can build the next version from scratch. Ten modules, explainable AI, a real BigCommerce integration, and a delivery model built for weeks rather than quarters.
Prepared by MageTech Solutions · Platform mts-bigcommerce-intelligence · Next.js 15 · NestJS 11 · PostgreSQL 16 · Redis 7 · Prisma 6 · BullMQ 5 · TypeScript 5
MTS BigCommerce AI Commerce Intelligence is a finished product, not a proposal. You can log in today and use every module. This document explains what is inside it, how we show it to you, what you receive, and everything else we can build, maintain or create from scratch.
A multi-tenant analytics and decision platform for BigCommerce merchants: sales, customers, products, inventory and marketing in one warehouse, with explainable AI insights, threshold alerts, an AI copilot, and scheduled reports — served through a role-aware web application.
A deployed, documented product with your store connected, your team trained, your roles configured, your data verified — plus the source code, environment documentation, CI pipeline and support terms. Full deliverables in section 7.
Extend this platform with custom modules, build a completely new product from scratch, modernise a legacy system, integrate other platforms, white-label it under your brand, or run it for you under a managed service. Section 9 lists every service line.
Most vendors sell a dashboard, a report, or a model. We hand over a working commerce platform, the engineering capability to extend it indefinitely, and a delivery model where a working version is in your hands in weeks — with the code and the knowledge to make it yours.
MageTech Solutions is a software product engineering company. We take ideas from discovery to a supported, running product — and we stay accountable for it afterwards.
We could describe our capability. It is faster to show it. This application is the reference implementation of how we work: structured, opinionated, complete and running.
Every figure on this screen is computed from the warehouse by the same services that run in production — the demo is the product, not a mock-up.
BigCommerce merchants have the data but not the answers. The control panel reports what happened; nobody joins sales, customers, stock and marketing into one decision view.
| Where it hurts today | What it costs | What this platform does |
|---|---|---|
| Sales live in the control panel, stock in a grid, marketing in ad tools | Decisions made on a single slice of the business | One warehouse, one joined view across all ten modules |
| Stockouts and churn discovered after the revenue is gone | Lost margin and repeat customers | Daily insight engine, coverage maths, at-risk segments, threshold alerts |
| Manual weekly exports rebuilt by hand | Hours of labour, inconsistent numbers, no audit trail | Parameterised CSV reports, scheduled delivery, stored history |
| Generic AI tools with no access to store data | Confident answers with invented numbers | Deterministic rules first, optional LLM explanation second, every figure traceable |
| Merchant data leaving the business for third-party AI | Privacy exposure and compliance risk | AI_PROVIDER=rule runs fully local; strict privacy mode blocks context sharing |
| Nobody can answer "who changed this, and when?" | Trust and audit failures | Role-based access, tenant isolation, and a full audit trail with actor and IP |
One number for revenue, orders, margin risk and customer value — without waiting a week for a report.
Coverage, low-stock and reorder signals early enough to act on, per variant and per location.
Campaign spend, attributed revenue and ROAS shown against store-wide performance, not in isolation.
Exports that reconcile, roles that restrict access, and an audit trail that satisfies a question months later.
Ten modules, one design language, one data layer, one permission model. Each module has its own accent colour, so the interface tells you where you are before you read a word.
Revenue, orders, AOV, customers, low-stock and open-alert counters; trends; channel mix; top products; latest AI insights; recent activity. What the owner opens first every morning.
Pipeline by status, revenue trend against the previous period, channel contribution, AOV movement, and order-level detail with customer context.
VIP, high-value, at-risk and inactive segments with counts and value at stake; lifetime value distribution; win-back targeting.
Performance by revenue and units, status mix, category and brand breakdowns, price and margin view, and a product health score for assortment decisions.
Days-remaining coverage, low-stock and out-of-stock queues, reorder points, stock trends and valuation, with variant-level drill-down.
Spend, attributed revenue, ROAS, CTR and conversion per campaign, channel efficiency comparison, and revenue trend in campaign context.
Generated insights with evidence and severity, prioritised recommendations, full history, and the AI copilot for natural-language questions.
Sales, inventory, customer and full-dataset reports, parameterised date ranges, CSV download, scheduled delivery and generation history.
Rules for revenue drops, order anomalies, AOV shifts, low stock and churn; triggered alerts with metric, value and threshold; in-app notification feed.
BigCommerce connection and sync history, team and roles, subscription and plan, AI provider configuration, notification preferences, audit log.
OAuth install or API-account connect; paged, resumable sync of categories, brands, products, customers, orders, inventory and settings; signed webhooks; visible checkpoints.
Shared metric computation — revenue, orders, AOV, growth, channel mix, stock cover, customer aggregates — used identically by the API, the insight engine and reports.
Five job queues and five schedules: dispatch syncs, nightly full sync, daily insights, alert sweeps and report delivery. Nothing heavy runs on a user request.
Seven roles, 25 permissions, tenant-scoped data access, hashed session tokens, encrypted integration secrets, and an audit trail of every administrative action.
One platform, five operating personas, each with a scoped view. This is what makes the product safe to give to a whole team on day one.
Anyone can show a chart. Showing the same screen to two people and letting them see different things — with server-side enforcement, not just hidden buttons — is how we prove the platform is enterprise-ready. In the demo we sign in as two different roles and let you notice the difference.
A demo is a sales tool. Ours is a working product walkthrough with a script, so you see the right thing in the right order — and you can judge the software rather than our enthusiasm.
admin@magetech.demo — we log in live, including the role switch, so you see authentication and access control rather than being told about them.admin@magetech.demoFull access. Used for the main walkthrough, insight generation and report export.The exact sequence we run, what we say, and the proof point each screen is there to prove. Use it to rehearse, or to run your own demo of the platform.
| # | Screen | What we do | What we say | Proof point |
|---|---|---|---|---|
| 1 | Marketing site & login | Show the public site, then log in live | "This is not a mock-up. You are about to use the same system we deploy for merchants." | Branding, product credibility, real authentication |
| 2 | Dashboard | Read the counters, switch the period, point out the insight feed | "Sixty seconds: where the business is, and what changed." | Speed to answer, and AI already summarised |
| 3 | Sales | Revenue trend, channel mix, order pipeline, drill into an order | "Every figure is computed from your own orders, not a spreadsheet." | Data integrity and drill-down |
| 4 | Customers | Open the at-risk segment, show value at stake | "Here is the revenue you have already earned and are about to lose." | Retention is measurable and targeted |
| 5 | Inventory | Sort by days of cover, open the low-stock queue | "This is the screen that prevents a stockout." | Operations value, variant-level detail |
| 6 | Marketing | Compare campaign ROAS, find the weakest channel | "Now we can tell you which spend to move." | Attribution joined to real revenue |
| 7 | Insights | Generate insights on demand, read one finding with its evidence | "Every insight shows its evidence. That is what makes it trustworthy." | Explainable AI, not a black box |
| 8 | AI copilot | Ask a question the customer chooses; try a second | "Ask it anything about sales, stock, customers or campaigns. The answer is built from your metrics." | Grounded AI with usage metering |
| 9 | Role switch | Sign in as Viewer, revisit Customers and Settings | "Same product, less access — enforced on the server, not just hidden in the menu." | Enterprise readiness |
| 10 | Reports | Generate a sales report, download the CSV | "This is what your finance team will actually use." | Deliverable artefacts, not dashboards |
| 11 | Alerts | Show rules, open a triggered alert and the notification feed | "The system watches the thresholds so your team does not have to." | Automation and proactive operation |
| 12 | Settings → BigCommerce | Show connection state, run a sync, show entity progress | "This is the live integration. OAuth or API token, resumable, with checkpoints." | Real integration, not an import script |
| 13 | Settings → Team & audit | Show roles, permissions and the audit trail | "Governance is not an add-on; it is in the data model." | Procurement and compliance confidence |
| 14 | Close | Agree the pilot scope and the success criteria | "Here is what we would build for you, and when." | A decision, not a follow-up email |
We run the demo on the seeded workspace and, in the same session, show the connection flow: the install link, the OAuth authorisation screen, and the API-token path. The customer sees the integration is real, and we agree the pilot's data backfill plan before anything is promised.
For a merchant with a specific question — "how do you show me slow-moving SKUs?" or "can your alerts feed our Slack?" — we prepare the workspace the night before so the answer is on screen, live, during the call.
And the answers we give — because nothing should be a surprise after a sale.
Orders, customers, products, inventory, categories, brands and settings sync on a paged, resumable basis with per-entity checkpoints, so an interrupted run continues rather than restarts. For a first load we agree a backfill window with you; historical depth is a plan decision, and we can start with a fixed period and extend it.
All platform access is isolated in one client package with a rate-limit-aware request layer, typed payloads and retries. When BigCommerce changes a payload, the change is contained there, covered by a regression test, and released without touching your modules.
Yes. The warehouse model is ours, so we add the entities and the sync jobs. Most integrations follow the same seven-step path used for BigCommerce, which is why a second source is fast.
They reconcile, and proving it is part of implementation. We reconcile a sample of periods and channels against the control panel before go-live, and any mapping difference is documented rather than smoothed over.
No. The insight engine is deterministic: revenue, product, inventory, customer, segment, order and marketing detectors evaluate your own metrics against configured thresholds and produce findings with evidence and severity. A language model, when enabled, only helps explain them.
Not by default. With the default configuration the entire AI layer runs locally. If you enable an external provider, privacy mode governs what may be sent: strict blocks context sharing, balanced sends redacted summaries, connected permits full context — and per-tenant usage and spend are recorded.
The system falls back to the local rule engine. Insights and answers continue; you simply lose the optional natural-language polish.
Yes — thresholds for revenue, order, AOV movement, at-risk recency, low-stock and overstock days are configurable, as are alert rules. We set sensible defaults during implementation and tune them with you once you see real behaviour.
Every business record carries a tenant id, tenant context comes from the session rather than the request, and the database client injects that scope on every query. Report files are partitioned per tenant as well.
Passwords are bcrypt-hashed, session cookies are opaque with only a hash stored server-side, and integration secrets such as BigCommerce access tokens are encrypted at rest with AES-256-GCM.
Stated plainly: inbound rate limiting, security headers, an OAuth state parameter, a CSRF token and password reset flows are hardening items on the phase-2 list. We schedule them as part of go-live, and we will not expose a deployment to the internet without them.
Yes. Administrative and configuration actions are recorded with actor, action, entity, metadata and IP address, scoped per tenant and visible to permitted roles.
Either your team, with our documentation, training and a support agreement; or MageTech Solutions under a managed service with monitoring, upgrades and response commitments. Both paths use the same runbook.
A working pilot with your own data inside one to two weeks of kickoff, a production-ready deployment in three to six, and ongoing managed support thereafter. Section 12 sets out exactly why it is this fast.
The source code, the documentation, the environment definition and the CI pipeline. If you later choose a different provider, you can continue without us — and we would rather earn the renewal.
That is normal and expected. Custom modules, integrations, forecasts and white-labelling are all standard work for us; we scope them separately from the platform subscription.
What you physically receive when the project is done. Nothing is implied, nothing is "available on request" — this is the checklist we close against.
Nothing in this list is "phase two of the engagement" unless you agree it in writing. When we say delivered, the item above is running in your environment, documented, and understood by your team.
Three distinct kinds of work. Most vendors only sell the first one and subcontract the third. We do all three with the same team.
New analytics modules, reports, alert types, dashboard widgets, integration entities, custom roles, API endpoints and copilot tool calls — added into the same architecture and design system so the product stays coherent.
Bug fixes, dependency upgrades, platform API changes, performance work, monitoring, backups, incident response, security patches and quarterly health reviews.
Entirely new products and platforms: discovery, architecture, UX, engineering, delivery, and support — the same discipline that produced this platform.
| Ask | What it looks like in practice | Typical shape |
|---|---|---|
| "We already have a BI tool" | Export paths and a data dictionary feed their warehouse; we add the commerce-specific logic they cannot model | Integration, days to weeks |
| "Our catalogue logic is unusual" | Custom product health, bundle and variant analysis, pricing and margin models on top of the existing schema | Custom module |
| "Alerts must reach our team" | Email or chat delivery for the existing alert engine, with tenant-scoped preferences | Integration |
| "We run several stores" | Multi-store tenants, per-store scoping, consolidated reporting and consolidated AI | Platform extension |
| "Sell this under our brand" | White-label: logos, colours, domain, copy, and tenant-isolated tenants for their customers | Reseller programme |
| "Prove value first" | Time-boxed pilot on your data with agreed success criteria and a written go/no-go | Pilot |
| "Replace an old system" | Legacy audit, migration plan, strangler rollout, parallel run and decommissioning | Modernisation |
The full capability set behind the platform. Any of these can be engaged standalone or as part of a delivery.
Requirements workshops, process mapping, data and integration audits, target architecture, and a written scope with success criteria before a line of code is written.
Information architecture, wireframes, visual design, a documented design system, accessibility and responsive behaviour, and prototypes for stakeholder review.
Modern front-end engineering with server rendering, component systems, form and state management, and performance budgets.
REST and service design, authentication and authorisation, validation, background jobs, caching, error contracts and integration endpoints.
Schema design, migrations, backfills, metric layers, data quality checks, warehouse modelling, snapshot and aggregation strategy, and export pipelines.
App development, OAuth and token integrations, V2/V3 API clients, webhooks, catalogue and inventory sync, custom storefront or extension work, and app-listing readiness.
Insight engines, copilot tool-calling, privacy and budget governance, provider selection, evaluation of output quality, and alert and workflow automation.
Report definitions, scheduled delivery, export formats, executive dashboards, ad-hoc analysis support and finance-grade reconciliation.
Containerisation, CI/CD, infrastructure as code, environment management, database and queue operations, backups, monitoring and cost control.
Test strategy, unit and integration coverage, browser end-to-end suites, regression discipline, performance testing and release verification.
Threat modelling, authentication and session hardening, secret management, access reviews, audit readiness, dependency scanning and security documentation for procurement.
Monitoring, incident response, upgrades, patches, performance tuning and a named support contact with agreed response expectations.
Auditing old systems, incremental replacement behind a facade, data migration with verification, and decommissioning the parts that no longer earn their keep.
Rebranding, custom domains, tenant-isolated deployments for your customers, and packaging the platform for agencies or groups of stores.
Training and enablement — role-based sessions, recorded walkthroughs and an administrator guide, so adoption is not dependent on us. Documentation as a deliverable — architecture, API, data model, development and administration documents are part of every handover, not an afterthought.
We are a TypeScript-first engineering team working across the modern commerce stack. This is the toolset behind the platform you just saw — and the toolset we bring to your project.
| Discipline | Technologies we work in daily | How we apply it |
|---|---|---|
| Language | TypeScript 5 (strict), modern JavaScript, SQL | One language across front end, back end, worker and shared packages; types as the contract |
| Web front end | Next.js 15 App Router, React 19, Tailwind CSS 4, SVG data visualisation | Server rendering for speed and SEO, client components only where interaction demands it |
| Backend | NestJS 11, REST API design, guards and interceptors, Zod validation | Thin controllers, service-layer business rules, consistent error contracts |
| Data | PostgreSQL 16, Prisma 6, migrations, indexing, aggregation | Normalised multi-tenant schema, pre-aggregated snapshots, migrations reviewed as SQL |
| Queues & jobs | Redis 7, BullMQ 5, repeatable schedulers | All long-running work off the request path, with retries and job retention |
| Commerce platforms | BigCommerce APIs v2/v3, OAuth 2.0, API accounts, webhooks, catalogue and inventory models | Rate-limit-aware clients, resumable sync, canonical channel mapping, signed callbacks |
| AI engineering | Rule engines, retrieval and tool-calling patterns, OpenAI / Anthropic / Gemini adapters, usage metering | Deterministic findings first, models as an explainability layer, privacy modes and budgets |
| Platform & DevOps | Docker, GitHub Actions, Turborepo monorepos, environment-driven configuration | Dependency-ordered builds, CI gates, reproducible environments, no secrets in the repository |
| Quality | Playwright, ESLint 9, static type gates, load and performance testing | Three quality layers and a release gate that blocks on regressions |
| Design systems | Token architecture, component libraries, brand theming, accessible interaction patterns | Colour, type, spacing and motion defined once and reused everywhere |
Code that another competent team can pick up on day one. Architecture that survives a second and third year of change. A product that looks deliberate, because it was.
One documented palette across the marketing site, the product and this document — the same rule we apply to your brand when we white-label.
An honest comparison, because you will be asked to choose between us, an agency, a freelancer, a SaaS vendor and a big consultancy.
| What matters to you | Generic agency | Freelancer / small team | SaaS vendor | Big consultancy | MageTech Solutions |
|---|---|---|---|---|---|
| Time to a working version | Slow, discovery-heavy | Can be quick, but narrow | Immediate but not yours | Very slow, heavy governance | Days for a pilot, weeks for production — because the core platform already exists |
| You own the code | Sometimes, reluctantly | Yes | No, subscription only | Yes, but as an asset you maintain | Yes, always — source, docs, CI, runbook |
| Customisation | Good | Good, if the skill exists | Only via extensions | Excellent and expensive | Good, on a platform architected for it |
| Speed of iteration after launch | Slow without budget | Depends on availability | Vendor roadmap only | Slow | Fast — the same team that built it |
| Maintenance and support | Extra contract | Ad hoc | Included, generic | Separate, expensive | Included options with agreed response terms |
| AI that is safe for your data | Unclear | Usually a public API | Vendor policy | Policy decks | Local-first rule engine, optional models, privacy modes and metering |
| Cost profile | High, unpredictable | Cheap, risky at scale | Recurring, no ownership | Highest | Platform subscription plus scoped work — no lock-in |
| Who does the work | Rotating staff | One or two people | N/A | Many juniors, few builders | Specialists who own the outcome end to end |
The platform is live today. Your pilot is a configuration and connection exercise, not a nine-month build. The hard parts are already solved and tested.
Code, documentation, infrastructure definition, CI. If we ever stop working together, you keep a product that still runs. That is a deliberate constraint on how we contract.
The people who architect and build the platform are the people who train your team and support it afterwards. Context is not lost at a handover.
Our AI is deterministic first. We would rather show you a finding with its evidence than a clever sentence that cannot be defended to your board.
Every engagement lands on a shared foundation: authentication, roles, design system, warehouse, jobs, CI. The second feature is cheaper than the first.
Known limits are in our documentation, not discovered in month three. Trust compounds faster than features.
Speed here is not heroics. It comes from starting on a finished platform, and from parallel work with a clear definition of done.
Timelines assume a single store or store group, an accessible API account, and a decision-maker available weekly. Multi-store, custom modules or white-label extend the plan — we will tell you before kickoff, not after.
Authentication, roles, modules, warehouse, jobs, AI engine, CI and deployment exist. We configure, we do not re-invent.
A realistic dataset ships with the product, so demos, development and testing start populated instead of empty.
Types are shared across the stack, so interfaces are agreed once and changes propagate visibly rather than being rediscovered.
Build, lint, type and browser tests run on every push, so quality is continuous rather than a final-week scramble.
Every new data source follows a proven path, so the second and third integrations are days, not months.
Configuration, backfill, documentation and training run alongside integration work instead of after it.
Integration and data verification — 18%
Configuration, roles and alert tuning — 14%
Backfill, reconciliation and hardening — 22%
Training, documentation and handover — 26%
Custom scope, if any — 20%
Note what is not in the list: months of groundwork. That work is already done, which is the entire point of buying into a platform rather than commissioning a rebuild.
When we add anything to a MageTech product, it follows the same path. Consistency is why delivery is fast and why the product stays coherent as it grows.
| Every release includes | Why it matters to you |
|---|---|
| Reviewed database migration | Your data history is safe and reversible |
| API contract kept in step | No surprise breakage for integrations |
| Type and lint gates green | No known-defect build reaches you |
| Browser end-to-end tests | The flows you actually use are verified automatically |
| Updated documentation | Your team is never working from stale instructions |
| Audit-visible change | Who changed what, when and why is recorded |
| Rollback path | Releases are reversible, not irreversible |
When we white-label or extend the platform for you, these rules become your product's rules. Your team inherits a codebase with a consistent shape — which is the difference between a product that grows and one that slowly becomes impossible to change.
Four ways to work with us. They can be combined — most customers start with a pilot inside Launch, then move to Managed.
A time-boxed engagement to stand up the platform for one merchant or store group and get the team using it.
You own the code, the documentation and the environment. If you later want to run it yourselves, you can — that is the point.
You use the platform; we carry operations, upgrades, monitoring and support with agreed response expectations.
Managed service means we operate your deployment, not that we own your business data. You keep access, exports and the right to leave with a complete handover at any time.
Fixed-scope feature or platform work, delivered in the same architecture, design system and quality gates as the product.
We always price the smallest thing that solves the problem. If an existing module or report with configuration does the job, that is what we will tell you.
For organisations with an internal engineering team that own the deployment and want senior guidance rather than a build contract.
"We built this ourselves — is it right?" Advisory engagements frequently start with a review of an existing internal system, then become a modernisation plan.
We are not only a product. MageTech Solutions is a BigCommerce development, customisation, integration, AI and managed-services partner — so a single fix, one new module and a full multi-year programme are all straightforward to buy. Seven ways to work with us, and every figure is a starting point, because the right price depends on scope, existing systems and integration complexity.
A specific requirement, bug fix, enhancement or small integration where the scope is not yet fixed.
One continuous senior resource on your team — part-time or full-time — for ongoing development and support.
Build or enhance one defined module or integration. The natural way to adopt the platform piece by piece.
A clearly defined deliverable with agreed scope, acceptance criteria and a fixed price.
Keep an existing application reliable with monitoring, fixes, small enhancements and technical reviews.
A complete MageTech engineering team — developer, backend, frontend, QA and lead — as an extension of your organisation.
Enterprise implementations, AI commerce projects, complex or multi-store integrations, ERP and CRM connectivity, data migration, custom BigCommerce applications and long-term product engineering. Scoped individually, with a written proposal before any work begins.
Every amount is a starting-from figure, not a fixed price. Final effort depends on your existing systems, the number of integrations, data volume, customisation and how much work is already reusable. We confirm effort, timeline, team and price in a written proposal before starting — we never invoice against an open-ended assumption.
For a specific requirement, bug fix, enhancement or small integration. Work is tracked, prioritised with you, and reported against the agreed estimate.
| Service | Recommended rate | Typical work |
|---|---|---|
| BigCommerce development | $25–$40 / hour | Storefront and app changes, catalogue, checkout, theme work |
| AI / commerce intelligence development | $35–$60 / hour | Insight detectors, copilots, recommendation and forecast logic |
| Integration / API development | $30–$50 / hour | BigCommerce, ERP, CRM, webhook and third-party connectivity |
| UI / frontend development | $25–$40 / hour | Design systems, components, data visualisation, accessibility |
| QA / testing | $20–$30 / hour | Functional, regression and browser end-to-end testing |
| Technical consultation | $40–$75 / hour | Architecture review, audits, technology selection, advice |
India-focused engagements: an equivalent starting range of ₹1,500–₹4,500 / hour, depending on skill level and complexity.
For customers who need a developer or technical resource continuously — typically an existing product team that needs additional BigCommerce, AI or integration expertise.
| Dedicated resource | Monthly starting from | Best suited to |
|---|---|---|
| Junior developer | $1,500 / month | Maintenance, small fixes, front-end work under guidance |
| Mid-level developer | $2,500 / month | Feature delivery across the stack, integration work |
| Senior developer | $3,500 / month | Architecture, complex features, code review, technical ownership |
| Senior AI / integration engineer | $4,000+ / month | Data platforms, AI systems, enterprise integrations |
Most clients who buy a dedicated resource already have a team and a roadmap; what they lack is specific depth. This adds one accountable senior engineer without a recruitment cycle, and it scales into a full team when the scope grows.
Instead of buying the whole platform, purchase the modules you need. Each is scoped, delivered and priced on its own, and every one runs on the same shared foundation — so nothing has to be rebuilt later.
| Module | Starting from | Module | Starting from |
|---|---|---|---|
| Commerce Dashboard | $1,500+ | AI Insights | $2,500+ |
| Sales Intelligence | $2,000+ | AI Recommendations | $2,500+ |
| Customer Intelligence | $2,000+ | Custom Reports | $1,000+ |
| Product Intelligence | $2,000+ | BigCommerce API Integration | $1,000+ |
| Inventory Intelligence | $2,000+ | External ERP / CRM Integration | $1,500+ |
| Marketing Intelligence | $2,000+ | Custom AI Integration | $2,500+ |
We use "starting from" deliberately: actual effort depends on scope, existing systems and integration complexity. Bundling modules is normally cheaper than buying them separately, and a full platform implementation is priced as a project — see below.
Start with the Dashboard and one intelligence module, prove the value on real data, then add the rest as budget allows. Because the data layer is shared from day one, each additional module is cheaper than the first.
For customers who want a clearly defined deliverable. Scope, acceptance criteria, timeline and price are agreed before work starts.
Depending on integrations, customisation, AI requirements, data volume and deployment model.
| Scope driver | Typical effect on effort |
|---|---|
| Number of stores | Each additional store adds connection, sync and reconciliation work |
| External integrations | ERP, CRM, marketplace and payment connectivity |
| Data history | Backfill depth and initial sync volume |
| Customisation | White-labelling, bespoke metrics, custom report formats |
| Deployment | Managed cloud, customer infrastructure or hybrid |
| AI requirements | Provider choice, privacy mode, volume and budget |
Storefront or BigCommerce app development, data migration from spreadsheets or legacy systems, custom reporting programmes, white-label platform builds, and performance or security remediation projects.
Ongoing support for an application we built or one we inherited. Priced monthly, cancellable, with defined response targets.
| Plan | Monthly | Response target | Included |
|---|---|---|---|
| Essential Support | $299 | Next business day | Bug fixes · basic technical support · minor configuration changes · monitoring · monthly maintenance |
| Business Support | $599 | Same business day | Everything in Essential · priority support · small enhancements · API and integration support · performance monitoring · monthly technical review |
| Enterprise Support | $1,499+ | Within hours, SLA-based | Priority and SLA-based support · production monitoring · advanced troubleshooting · integration support · security and technical reviews · continuous improvements · dedicated support contact |
BigCommerce stores and apps, the MTS intelligence platform, custom web applications, APIs and integrations we built — and inherited applications after a review.
A managed service where we operate and monitor the platform and you simply use the product; or a support-only arrangement on your own infrastructure.
Unrequested feature development. That is quoted separately as a module, project or hourly item, so support cost never hides development cost.
For larger customers who want MageTech Solutions to operate as an extended development team rather than a sequence of projects.
A typical team, scaled to the requirement:
Composition, seniority and hours are agreed per engagement. A single dedicated resource is available at a lower entry point — see model 02.
| Aspect | How it works |
|---|---|
| Delivery | Short iterations against a prioritised backlog you influence |
| Communication | Weekly demonstration, written decisions, one escalation path |
| Quality | The same pipeline standard applies to every change we make in your account |
| Documentation | Updated in the same change, not at the end of the engagement |
| Ownership | Your repository, your infrastructure, your cloud — we work in your estate |
| Exit | Source, documentation and configuration are yours throughout |
Most teams start with BigCommerce, backend and QA capacity, add a dedicated resource from model 02, and take AI and data engineering on demand as the intelligence layer grows.
We do not price only by hours, because that positions us as a staffing supplier rather than a technology partner. The right model follows the shape of the requirement.
A fix, an enhancement, a question, something undefined. Model 01 from $25/hour.
A defined module, integration or report with a known outcome. Models 03 and 04 from $1,000 and $5,000.
A full implementation with scope, milestones and acceptance criteria. Model 04 from $5,000, typically $10,000–$30,000+.
An ongoing roadmap needing steady capacity. Model 02 from $1,500/month or model 06 from $6,000/month.
Something already works and must keep working. Model 05 from $299/month.
Multi-store, ERP and CRM, migration, AI programmes, long-term product engineering. Model 07, scoped individually.
We recommend the model that fits the requirement, not the one that maximises revenue. A client who starts with a $1,500 module and a good experience comes back for the platform; a client over-committed to a large fixed scope does not.
Enterprise implementations, AI commerce projects, complex or multi-store integrations, ERP and CRM connectivity, custom BigCommerce applications, data migration and long-term product engineering all start with the same sentence: tell us what you need, and we will design the right engagement model.
The product does not end at go-live. This is what happens next, and what we commit to.
Every engagement has a documented escalation path with named contacts, a severity definition, and a response commitment agreed in writing before work starts. Nothing is implied — if it is not in the agreement, it is not promised.
This is an honest list: most delays in delivery projects are caused by access or decision latency, not engineering. Naming it up front is faster than discovering it late.
The commitments behind the platform, and the summary you can hand to your procurement or security review.
Strict types across every workspace, lint gates, a single committed migration, dependency-ordered builds, and browser end-to-end coverage of authentication, modules and integration flows.
Item prices snapshotted at order time, so historical revenue never shifts. Idempotent upserts keyed on external identifiers. Reconciliation against source figures before go-live.
Hashed passwords, opaque hashed session tokens, encrypted integration secrets, server-side permission enforcement, tenant-scoped data access, signed webhooks and a full audit trail.
Local-first AI with explicit privacy modes, per-tenant usage metering, and no merchant data leaving your infrastructure unless you explicitly enable it.
Health endpoints, structured logging, background job visibility with retries and retention, backup and restore drills, and documented rollback for every release.
Known limitations are documented in our own materials. Hardening items are named, scheduled and completed before public exposure — not discovered later.
| Control | Implementation in this platform | Status |
|---|---|---|
| Password storage | bcrypt hashing; plaintext never stored or logged | Live |
| Session management | Opaque token in an HttpOnly, SameSite cookie; only a hash stored | Live |
| Integration secrets | AES-256-GCM encryption at rest with a supplied key | Live |
| Authorisation | Permission guard on every sensitive endpoint; 51 enforced permissions | Live |
| Tenant isolation | Session-derived tenant scope injected at the database client | Live |
| Audit trail | Actor, action, entity, metadata and IP for administrative changes | Live |
| Webhook verification | Raw-body HMAC verification with event-hash idempotency | Live |
| Rate limiting and security headers | Hardening phase, scheduled before internet exposure | Phase 2 |
| OAuth state and CSRF token | Added with the PKCE and session-rotation work | Phase 2 |
| Password reset, MFA, email verification | Account lifecycle phase | Phase 2 |
We agree measurable success criteria before the pilot and report against them afterwards. These are the categories we hold ourselves to.
| Area | What success looks like | How we measure it |
|---|---|---|
| Time to insight | Minutes from question to answer, instead of a manual report cycle | Adoption and response-time feedback from the team using it weekly |
| Stock and service levels | Fewer avoidable stockouts on priority SKUs | Coverage alerts raised versus stockout events |
| Customer retention | At-risk customers identified before they lapse | Win-back campaign response from the at-risk segment |
| Marketing efficiency | Spend shifted toward campaigns that actually return revenue | ROAS movement on reallocated spend |
| Reporting effort | Finance cycles rebuilt by hand, replaced by exports | Hours spent on manual report preparation |
| Trust and governance | Every number traceable, every action audited, every role scoped | Reconciliation success, audit completeness, access reviews |
| Delivery certainty | Working version early, no surprises at handover | Milestone adherence against the agreed plan |
Customer references, case studies and a fuller security and compliance pack are available during evaluation, under the terms your procurement process requires. We would rather you verify us than take our word for it.
Every number we show you in a demo comes from the running product. Every capability we describe is either built, or labelled as roadmap. If it is not in the system, we will say so.
So a demo never goes wrong. Use this before you present the platform to a customer — or ask us to present it with you in the room.
If we join your call, we will prepare the workspace around the questions the customer raised in discovery, so the answers are on screen rather than promised. Tell us the dealer's industry, store size and the two questions they care about most.
Three ways forward, in increasing order of commitment. Pick whichever matches your risk appetite.
Thirty minutes on a seeded workspace, with your questions answered live. You keep access afterwards to explore on your own.
Best for: confirming the product is real and the answers match your business.
Time-boxed, with your store connected, your numbers verified, your team trained, and written success criteria agreed up front.
Best for: proving value with your own figures before committing.
Full implementation, custom development, or a managed service where we run the platform and you run your business.
Best for: committing to the platform and building on it.
Contact MageTech Solutions Book a live demo
MageTech Solutions · MTS BigCommerce AI Commerce Intelligence · www.magetechsol.com
Quick reference for the platform, the demo environment and this document.
http://localhost:3000 in developmenthttp://localhost:3001/apiadmin@magetech.demo · MTS-demo-2026!| Entity | Count | Entity | Count |
|---|---|---|---|
| Products | 1,284 | Customers | 12,450 |
| Orders | 8,420 | Order items | 21,160 |
| Revenue | $284.5K | Daily snapshots | 366 |
| Campaigns | 15 | Segments | 6 |
| Insights | 7 | Recommendations | 2 |
| Alert rules | 5 | Roles / permissions | 7 / 25 |
| Item | Detail |
|---|---|
| Modules | Dashboard, Sales, Customers, Products, Inventory, Marketing, Insights, Reports, Alerts, Settings |
| API | 64 endpoints, 18 controllers, session and permission guarded |
| Data | PostgreSQL 16 with 32 models and 30 enums, tenant-scoped |
| Jobs | 5 queues, 5 schedules, retries with exponential backoff |
| BigCommerce | OAuth 2.0, API-account token, signed webhooks, 7 sync entities |
| AI | Rule engine with 7 detectors, optional LLM copilot, 3 privacy modes |
| Exports | 4 report types, CSV, tenant-partitioned, schedulable |
| Quality | Strict types, lint, 10 end-to-end tests, CI on every push |
docs/