AI Site Biz
AI Site Biz build series
AI website companion for owner-operated service businesses.
001
Index - an applied AI product practice
JetCalls builds applied AI products with real infrastructure behind them.
We design, build, and operate product-grade software across AI websites, real-estate intelligence, AI-guided development, governed dashboards, and hosted business agents.
Applied AI
Productized workflows
Website generation, AI analysis, dashboard specs, and business-agent tasks.AI-guided coding
Feature-by-feature delivery
Analyst, Developer, Tester, and Preview agents guide small changes from question to live URL.JetCalls
Build the product and the operating system around it.
Strategy, engineering, data, deployment, monitoring, and durable context.Data platforms
Decision systems
ZIP-level housing data, finance connectors, reports, and validated dashboards.Hosted agents
Tenant-aware autonomy
Isolated memory, schedules, A2A access, access metadata, and runtime controls.Portfolio scope
5 products
AI productization, data platforms, guided coding, dashboards, and hosted agents.
002
AI websites
AI Site Biz turns owner briefs, old sites, and WhatsApp-style change requests into managed static websites.
Real-estate intelligence
ReSmart.ai analyzes ZIP markets, listings, affordability, reports, and AI-agent access to housing data.
Freeform coding
Freeform Code proves AI-guided software delivery where agents ask questions, scope one feature, build, test, and return a live preview.
Governed dashboards
AI Dashboard converts natural language into validated dashboard specs from approved data sources.
Agent infrastructure
Hermes Hub hosts tenant-scoped business agents with schedules, A2A, runtime inspection, and isolation.
003
Blog - product development stories
The JetCalls blog is about how we build.
The products are proof points. The blog explains the process behind them: what we chose, what we avoided, what we shipped, and what each product taught the team.
ReSmart.ai
ReSmart.ai build series
ZIP-level real-estate intelligence and AI second opinions.
- ReSmart.ai: Turning Property Search Into Guided Questions 2026-02-25
- ReSmart.ai: Making Property Analysis Shareable 2026-02-28
- ReSmart.ai: Adding Compliance and Status-Aware Chat 2026-03-08
- ReSmart.ai: Building Market Report Pages for Search 2026-04-10
- Building ReSmart.ai: Vertical Real Estate Intelligence 2026-04-22
Freeform Code
Freeform Code build series
Inversion-of-control software development led by AI agents.
AI Dashboard
AI Dashboard build series
Governed prompt-to-dashboard generation for approved data.
- AI Dashboard: From Investor Mockup to Product Plan 2026-04-26
- Building AI Dashboard: Governed Prompt-to-Dashboard Generation 2026-04-28
- AI Dashboard: Saving Dashboards and Adding Real Estate Data 2026-05-04
- AI Dashboard: Connecting Finance Data and Options Signals 2026-05-07
- AI Dashboard: From Shared Demo to User Workspaces 2026-05-18
- AI Dashboard: Making Market Summaries Easier to Scan 2026-05-20
Hermes Hub
Hermes Hub build series
Hosted business agents with isolated memory, schedules, and controls.
- Hermes Hub: Building the Tenant Runtime Foundation 2026-05-02
- Hermes Hub: A2A Discovery and Scheduled Wakeups 2026-05-04
- Building Hermes Hub: Hosted Business Agents 2026-05-05
- Hermes Hub: Runtime Policy, Usage, and Business Metrics 2026-05-11
- Hermes Hub: Admin UI and Runtime Lifecycle Control 2026-05-14
- Hermes Hub: Making Hosted Agents Easier to Use 2026-05-25
004
Services - five
Five service lines, one delivery model.
JetCalls is useful when a project needs both product imagination and infrastructure discipline. We do not stop at a landing page, prototype, or prompt; we ship the system around it.
Strategy
Product framing, technical due diligence, SEO and market positioning, implementation plans, and modernization roadmaps.
Build
Full-stack product engineering across static sites, web apps, APIs, data pipelines, billing, tests, and deployment.
Platform
Cloudflare, Firebase, NestJS, Hugo, worker runtimes, tenant models, release flows, observability, and operational tooling.
Applied AI
LLM workflows, validated specs, AI assistants, website generation, search/data enrichment, evaluations, and agent skills.
Operate
Managed improvements, production checks, context wikis, cron jobs, quality gates, and maintenance loops that keep products alive.
005
Products - five product proofs
The products show what JetCalls can actually build.
These are not abstract service claims. Each product demonstrates a different capability: SMB AI productization, real-estate data intelligence, AI-guided development, governed analytics, and hosted AI agents.
AI Site Biz
AI website companion for owner-operated service businesses.
AI Site Biz builds and maintains fast small-business websites from owner briefs, old sites, brochures, photos, and chat-style update requests. The important idea is not generic AI design; it is a practical website helper that drafts changes and waits for owner approval.
- Hugo site generation, content planning, image jobs, quality gates, and Cloudflare deployment are implemented.
- Strong SEO signal around AI website builders, contractor website design, and website maintenance cost.
- Public wording should avoid guarantees like ADA compliant, fastest, or guaranteed ranking.
Product proof - AI productization for trades and local services
ReSmart.ai
ZIP-level real-estate intelligence and AI second opinions.
ReSmart.ai helps buyers, sellers, renters, landlords, investors, agents, and AI agents evaluate US housing markets, listings, affordability, timing, and risk from public market data plus guided AI analysis.
- Frontend, backend, ingestion, reports, listings, guest chat, API keys, MCP tools, and monitoring are evidenced in the product repos.
- Best SEO routes are cap-rate calculators, real-estate market reports, analytics, and AI real-estate assistant pages.
- Public claims need care around forecasts, school/crime data, and external data licensing.
Product proof - vertical AI and data intelligence
Freeform Code
Inversion-of-control software development led by AI agents.
Freeform Code changes the shape of coding work. Instead of a human planning a whole product and instructing an agent step by step, an Analyst agent asks questions, scopes one small feature, and drives the work through Developer, Tester, and Preview agents until there is a live URL to inspect.
- Repo evidence includes A2A agent interfaces, per-agent sessions, push notifications, debugging slot, Telegram and Teams bridges, pinned specs, and preview routing.
- The workflow focuses on one topic, one small feature, one turn of progress, with visible handoffs and a durable spec mirrored into the repo.
- Client usage exists but is not named publicly; public proof should stay centered on the repo, internal usage, and live preview capability.
Product proof - agent-guided software development
Anton -> Analyst
Add Google sign-in. Normal email login keeps working.
Analyst
Three questions, then I will write the pinned spec and ask for approval.
Analyst
Asks, scopes, writes acceptance criteria.
Developer
Branches, edits, commits, and hands off.
Tester
Runs the build and real browser checks.
Preview
Starts the service and posts the URL.
Live result
https://add-oauth.techflow.jetcalls.com/
AI Dashboard
Governed prompt-to-dashboard generation for approved data.
AI Dashboard converts plain-English requests into validated dashboard specifications, then renders persistent charts, KPIs, tables, and summaries from approved finance and real-estate connectors.
- The core loop exists: Firebase auth, generation, CRUD, version history, validated Zod specs, connectors, and React rendering.
- SEO demand is smaller but high-value around AI business intelligence, AI dashboard builder, and AI dashboard generator.
- Avoid claiming universal BI replacement, any data source, full MCP marketplace, or enterprise readiness until proven.
Product proof - governed AI analytics
Hermes Hub
Hosted business agents with isolated memory, schedules, and controls.
Hermes Hub operates tenant-scoped Hermes Agent runtimes: each business gets its own workspace, memory, schedules, access metadata, A2A boundary, and runtime controls. It is the infrastructure proof behind JetCalls' agent work.
- Evidence includes tenant CRUD, runtime inspection, A2A agent cards, cron, central releases, metrics, HubChat, and business-agent scaffolds.
- SEO needs two tracks: SMB business-assistant pages and technical A2A/control-plane pages.
- Avoid fully autonomous business, enterprise scale, and marketplace-first claims until those paths are proven.
Product proof - agent infrastructure
tenant_roomcord_website
isolated workspace, cron schedule, A2A card, and runtime socket.
business companion
weekly review loop, wiki memory, website skills, and guarded initiative.
control plane
tenant lifecycle, releases, env metadata, metrics, and inspections.
A2A boundary
external discovery and messaging without exposing private runtime control.
006
Proof - implementation-backed positioning
What the portfolio says about JetCalls.
The clearest company story is that JetCalls builds applied AI products with the operating layer included: product UX, data, agent workflows, deployment, tests, and durable context.
33,100
Monthly US searches for "ai website builder"
AI Site Biz can earn attention through broad AI-site-builder demand, then convert through trade-specific pages.9,900
Monthly US searches for "cap rate calculator"
ReSmart.ai should use calculator and market-report pages as entry points into ZIP intelligence.4
Agent roles in the Freeform Code delivery loop
Analyst scopes and asks; Developer builds; Tester verifies; Preview launches a real URL for review.1,000
Monthly US searches for "agent to agent protocol"
Hermes Hub can speak separately to technical A2A buyers and small-business agent buyers.007
Tell us what you are trying to build.
Bring the business problem, the messy repo, the product idea, or the infrastructure gap. JetCalls can help turn it into a working system and the public story around it.