Bounded agent systems grouped by the task they can safely own.
Voice, chat, SMS, email, reminders, drafting, routing, and human review.
AI Agents HMX Zone.
HMX Zone
ai agent service
Agent system objects
Voice, chat, SMS, email, reminders, drafting, routing, and human review.
Medium system
A time-aware agent that covers nights and weekends: it answers, captures the request, books where it can, and decides what truly needs to wake a human versus what waits for morning. Keeps speed-to-lead alive off-hours without paying for an overnight desk.
timeline1-2 weeks
fallbackDefined
High system
An operations view built specifically for conversational agents: containment rate, escalation rate, average handle time, latency, opt-outs, booking conversion, and cost per resolved conversation. The single pane that tells you whether the agent is actually working, not just running.
timeline1-3 weeks
fallbackDefined
Medium system
The tool-call layer that lets a voice or chat agent actually book a meeting: check real availability, hold a slot, confirm timezone, write the event, and send confirmation. It connects the conversation to a real calendar so 'I booked you for Tuesday' is true, not a hallucinated promise.
timeline4-9 days
fallbackDefined
Medium system
A post-call step that turns a voice or video transcript into a structured summary: outcome, key facts, objections, next action, and owner. It writes the recap straight onto the lead record so the next person picks up with full context instead of relistening to the call.
timeline3-6 days
fallbackDefined
Medium system
The write-back layer that lands every agent conversation onto the right CRM record: matched contact, structured summary, outcome, next step, and recording link, with deduplication and idempotency. Makes the agent a first-class contributor to the CRM instead of a side channel that loses history.
timeline4-9 days
fallbackDefined
Medium system
An agent that reads incoming email, classifies it (sales lead, support, billing, spam, escalation), extracts the key ask, applies a label/route, and drafts a suggested reply for the right owner. It sorts and prepares the inbox; a human still owns the send on anything sensitive.
timeline1-2 weeks
fallbackDefined
Low system
The designed 'I don't know' path: a small library of safe, branded responses the agent uses when it lacks a grounded answer, hits a blocked topic, or fails a tool call, each paired with a useful next step. Replaces hallucination and dead silence with a graceful exit.
timeline2-4 days
fallbackDefined
High system
The handoff machinery that moves a conversation from agent to person cleanly: detect the escalation trigger, package the context, route to an available owner by channel, and keep the customer informed. Turns 'let me get someone' into a real, tracked transfer instead of a dead end.
timeline1-2 weeks
fallbackDefined
High system
A retrieval layer that keeps an agent answering only from approved, current source material. It grounds responses in a curated knowledge base, cites what it used, and refuses or escalates when the answer is not covered, so the agent stops improvising facts about pricing, policy, or availability.
timeline1-2 weeks
fallbackDefined
Medium system
A structured measurement pass that breaks a voice agent's response delay into its parts (STT, LLM, TTS, network, turn detection) and checks each against targets so calls feel like a conversation, not a walkie-talkie. The diagnostic that finds where the lag actually lives.
timeline3-6 days
fallbackDefined
Medium system
An agent step that turns a raw conversation (call, chat, or SMS) into a structured fit-and-intent score with a short rationale, so leads route to the right owner by quality instead of arrival order. Reads the conversation the way a closer would and writes the verdict to the CRM.
timeline4-8 days
fallbackDefined
Medium system
A compliance-grade layer that detects and honors opt-outs across voice, SMS, and email: recognize STOP/unsubscribe/'do not call', suppress the contact everywhere immediately, confirm, and keep a permanent record. The non-negotiable plumbing that keeps an outreach agent legal and trusted.
timeline4-8 days
fallbackDefined
High system
A CI-style test suite that pins agent behavior so a prompt or model change cannot silently break what already worked. Every fixed scenario asserts the right outcome (correct extraction, correct refusal, correct escalation), and the suite runs before any change ships.
timeline1-2 weeks
fallbackDefined
High system
A repeatable test rig that runs the same scripted scenarios through Vapi, Retell, and Bland and scores them side by side on latency, interruption handling, transcription accuracy, task completion, and cost. Turns provider selection into evidence instead of a vendor pitch.
timeline1-2 weeks
fallbackDefined
High system
A defined boundary for topics the agent must not handle on its own — medical, legal, financial advice, threats, or anything regulated — where it safely declines, gives an approved response, and routes to a qualified human. Keeps the agent from giving advice it is not allowed to give.
timeline4-9 days
fallbackDefined
Medium system
A two-way SMS agent that reads inbound texts, drafts a context-aware reply using the contact's CRM history, and either sends within guardrails or queues the draft for one-tap human approval. Handles the texting cadence (quiet hours, length, STOP) that generic chatbots get wrong.
timeline4-8 days
fallbackDefined
High system
An explicit allow-list defining exactly which actions an agent can take, with what inputs, and which require confirmation or a human. It bounds the blast radius so an agent can read freely but cannot refund, delete, or send anything irreversible on its own.
timeline4-9 days
fallbackDefined
Medium system
A weekly quality loop that samples real agent conversations, scores them against a rubric (accuracy, escalation correctness, tone, task completion), and feeds the failures back into prompt and guardrail fixes. The mechanism that keeps a live agent from quietly drifting after launch.
timeline1-2 weeks
fallbackDefined
High system
An inbound or outbound phone agent that greets the caller, asks a short fixed qualification script (budget, timeline, location, intent), confirms key answers back, and writes a structured fit score to the CRM. Built provider-aware so the same script runs on Vapi, Retell, or Bland depending on volume, latency, and call-control needs.
timeline1-2 weeks
fallbackDefined
Medium system
A web-chat agent that handles first-touch intake: it collects name, contact, and reason for reaching out, answers a handful of grounded FAQs, and either books a slot or opens a CRM lead. Designed to qualify and route rather than to free-roam, with a tight scope and a fast handoff to a person.
timeline3-7 days
fallbackDefined