Blog banner

From our blog — ideas worth reading

  • 3 min read

    Spaceplace Is Heading to TechCrunch Disrupt 2026

    From October 13 to 15, Spaceplace will be in San Francisco for TechCrunch Disrupt 2026. We are one of nine startups selected for the Ukrainian delegation. We are coming with one goal: to show how AI agents work inside real business operations. What is TechCrunch Disrupt? Disrupt is one of the best-known startup events in tech. Every year it brings founders, investors and industry leaders together in San Francisco, and this year it takes place at Moscone West. The organizers expect more than 10,000 attendees and over 20,000 professional connections. Startup Battlefield, the event’s pitch competition, counts alumni who have raised a combined $32 billion. Some familiar names started here Many products people use every day first met the public on this stage: Mint launched at TechCrunch40 in 2007 and won the competition. Two years later, Intuit acquired it for $170 million. Dropbox launched publicly at TechCrunch50 in 2008. Trello launched at TechCrunch Disrupt San Francisco in 2011. From early-stage products to companies used by millions, Disrupt has a long record of putting new ideas in front of the right people. Part of the Ukrainian delegation Spaceplace is one of nine Ukrainian startups selected for Disrupt 2026. The delegation is organized by the Ministry of Digital Transformation of Ukraine and the Ukrainian Startup Fund, with support from the Ukraine-Moldova American Enterprise Fund and YEP Accelerator. We are proud to stand next to teams building in robotics, cybersecurity, agritech and defense tech. Ukraine first brought its own pavilion to Disrupt in 2022, and the delegation keeps growing in ambition. Ukrainian tech keeps building, shipping and competing on the global stage. What we are bringing to San Francisco Spaceplace is an AI agent operating platform for building, managing and scaling AI agents across an organization. It is one environment where people, agents and business context work together. At Disrupt, we will show: Agents that do the work, not just answer. They reply to customers, qualify leads, book meetings and update records in the CRM, ERP, telephony and messengers a company already runs. Shared business context. Every agent draws on the same company memory, knowledge and rules, so a conversation that starts in one channel continues in another. Human control by design. Sensitive steps wait for a person’s approval, and every action is logged: who or what acted, when and on which rule. Real workflows, live. From an inquiry to a booked viewing, from a question to a resolved case, from a call to a scored conversation. Who we want to meet Enterprise teams with one high-friction workflow they would like to hand to an agent Partners and integrators bringing AI into their clients’ operations Investors focused on agentic AI for enterprise operations Fellow founders, to compare notes on building AI products from Ukraine See you in San Francisco If you are at Disrupt on October 13–15, come and say hello. Want to book time with our team in advance? Get in touch and we will set up a meeting. Follow Spaceplace on LinkedIn for live updates from San Francisco.

    • Product Updates
  • 4 min read

    Communication Intelligence in Practice

    Most companies do not have a communication problem. They have a connection problem. The buyer who wrote on Instagram, called the sales office and emailed a passport scan is three different people in three different tools. Communication intelligence fixes that. It treats every conversation as part of one relationship, understands what is being said and acts on it. This article shows what that looks like in practice. Three layers, not one chatbot Communication intelligence works on three layers. Remove any one of them and you are back to a chatbot or a reporting tool. Connect. Phone, messengers, email and web chat feed one customer record and one shared memory. Every agent and every employee sees the same history. Understand. Each conversation is turned into structured meaning: intent, budget, timeline, objections, missed opportunities. Calls are transcribed and evaluated, not just recorded. Act. The next step happens in your real systems. The agent replies, books, updates the CRM, schedules a follow-up or routes the case to the right person. Without the first layer, every channel starts from zero. Without the third, you get insights that nobody acts on. One customer, three roles Spaceplace’s Communication Intelligence pack follows the customer from lead to buyer to resident. Here is an illustrative journey for one person, Maria. Maria the lead. On a Tuesday at 22:15, Maria asks on the website chat about two-bedroom apartments. The agent shows units that are actually available, asks about budget and timeline one question at a time, and books a video call for Thursday. Her CRM record already holds the brief. Maria the buyer. Three weeks later, she calls the sales office about installment terms. The operator sees the full history before saying hello. After the call, the transcript is summarized, the objection is detected (the monthly payment feels too high), and the CRM gets a next action: send the longer payment plan. The follow-up reaches her on WhatsApp the same afternoon. Maria the resident. A year after handover, she reports a leaking tap in the messenger she always uses. The request becomes a service ticket, a contractor is assigned, and she gets status updates on the same channel. The property team knows exactly who she is. Three roles, one memory, no repeated questions. That continuity is what customers notice, even if they never hear the term “communication intelligence”. Inside the call center The phone is still where many high-value conversations happen. Spaceplace’s Call Center Intelligence pack transcribes and evaluates every call, instead of the small sample a supervisor has time to hear. It detects objections and missed opportunities, writes the summary and triggers the next action in the CRM. This changes how teams improve. Supervisors coach operators on specific moments, not on general impressions. Recurring objections, such as concerns about delivery dates or payment terms, become visible across all calls and can feed back into scripts, pricing and marketing. What changes for each role Sales manager. Before: rebuilds the buyer’s history from scattered chats. After: opens a briefed record with the full thread. Call center lead. Before: reviews a handful of calls per operator. After: sees every call scored, with objections flagged. Property manager. Before: requests arrive by phone, email and messenger. After: one ticket per request, with status and owner. Buyer or resident. Before: repeats the story on every channel. After: is recognized wherever they write or call. Guardrails that make it trustworthy Connecting every conversation only works if access stays controlled. In Spaceplace, each agent and each person sees only the sources their role permits. Sensitive steps wait for human approval, and a full audit trail records who or what acted, when and on which rule. Data stays in the company’s own tenant. Metrics worth tracking First response time across all channels Share of conversations linked to a CRM record Most frequent objections, by type and by project Share of quiet leads re-engaged by follow-up Time from a resident request to resolution Where to start Start with the channel that leaks the most, often messengers or after-hours calls. Connect it, let the agent handle routine conversations, and keep people on the exceptions. When the first channel runs cleanly, add the next one to the same memory. See communication intelligence on your own conversations. Book a live demo and bring one real customer journey.

    • AI Agents
    • Guides & Tutorials
  • 3 min read

    The Superpowers of Real Estate AI Agents

    Real estate sales run on speed, trust and follow-through. Buyers message late at night, compare several developers at once and expect straight answers about price and payment plans. Your best agent can do all of that, just not for every lead, on every channel, at every hour. AI agents can. Here are the eight superpowers that make the difference, including the one that matters most: knowing when to step aside. 1. They never keep a buyer waiting The office closes at 19:00 and a buyer writes at 23:40. The AI agent qualifies them, matches real units and books a viewing while the team sleeps, so the lead is on a manager’s desk at 9:00. Spaceplace reports an average response time of under 3 seconds, on any channel, at any hour. 2. They know your inventory by heart The agent reads your live inventory, not last week’s spreadsheet. It quotes only what is actually available, so a unit sold this morning never shows up in tonight’s offer. Property cards, photos and tours appear right in the chat, with no redirects. 3. They do the math on the spot Installment plans and mortgage programs are where many buyers get stuck. The agent explains program terms, works through eligibility questions and calculates the monthly payment in plain numbers. A question that used to wait for a callback gets an answer in seconds. 4. They speak every buyer’s language, on every channel Buyers choose the channel: website chat, Instagram, WhatsApp, Telegram, Viber or email. The agent replies where they wrote, in the language they wrote in. Every conversation ends up in one CRM record, deduplicated, instead of on one manager’s personal phone. 5. They qualify without pressure Nobody enjoys a questionnaire. The agent asks one question at a time: budget, number of rooms, area, timeline, financing. It scores the lead as it goes, so your closers receive only buyers who are ready, with intent already captured. 6. They own the calendar The agent offers real slots, books the viewing or video call, and confirms it with the buyer. It sends reminders 24 hours and 2 hours ahead. When someone still misses the slot, it re-books them without anyone chasing. 7. They never forget a buyer Every conversation, preference and objection is stored in the company’s operational memory. The next conversation starts where the last one ended, even if a different agent or a different manager picks it up. When a manager leaves the company, the relationship stays. 8. They know when to hand over The most important superpower is restraint. The agent acts when it is sure and asks when it is not. Sensitive or irreversible steps, such as a price exception, stop for a person with the evidence and a recommendation attached, and every action is logged. The team behind the superpowers In Spaceplace, these abilities are split across specialized agents that share one memory, so every handoff carries the full context. Lead Qualifier — replies in seconds, captures the brief, scores the lead, writes a clean CRM record. Stage: inquiry and qualification. Property Consultant — matches live units to the brief, shows property cards, runs installment math. Stage: consultation. Booking Specialist — books viewings and video calls, sends reminders, recovers no-shows. Stage: viewing. Follow-up Manager — re-engages quiet leads on a timed sequence, flags warm leads going cold. Stage: conversion. Document Assistant — prepares leases and contracts, runs KYC checks, keeps an audit trail. Stage: paperwork and signing. Beyond the sale The superpowers do not switch off at signing. After handover, the same platform coordinates resident requests, maintenance, service tickets, payment reminders and contractor handoffs. The buyer becomes a resident, and the company still remembers who they are. What it adds up to For real estate teams, Spaceplace reports under 3 seconds average response time, 45% more qualified leads and 3 times more viewings. Your people keep the work that needs a human: building trust, negotiating and closing. See the agents work on your project. Book a live demo and bring one real inquiry flow.

    • AI Agents
    • Industry Insights
  • 4 min read

    How Intelligent Automation Improves Sales

    Most deals are not lost at the negotiating table. They are lost earlier, in the gaps nobody owns: the message that arrived at 23:40, the follow-up that never went out, the viewing nobody confirmed. Intelligent automation closes those gaps, and that is where most of its impact on revenue comes from. Rules versus intelligence Classic sales automation runs on rules. If a form is submitted, send template B; if there is no reply in three days, send template C. It is predictable, but it cannot read a question, check a price or decide what should happen next. Intelligent automation uses AI agents that understand intent and work inside your systems. An agent reads the buyer’s message, checks live inventory, answers with real numbers, writes the record to your CRM and books the meeting. When it is not sure, it hands the case to a person with the full context attached. The difference matters because sales conversations rarely follow a script. A rule can send a message. An agent can hold a conversation. Where automation moves the numbers Speed to lead. In a Harvard Business Review audit of 2,241 US companies, the average first response to an online lead took 42 hours. The MIT/InsideSales Lead Response Management study found that leads contacted within 5 minutes are 21 times more likely to qualify than leads contacted after 30. An AI agent replies in seconds, at any hour, so response time stops depending on who is awake. Qualification without the interrogation. The agent asks one question at a time: budget, timeline, need, decision process. Answers are scored and saved as structured CRM fields, not buried in a chat log. Ready buyers move forward; the rest still get a useful answer. One record per buyer, whatever the channel. Leads write through the website, Instagram, WhatsApp, Telegram and email, and many of those threads end up on one manager’s phone. An agent turns every lead into a single deduplicated CRM record with the full conversation attached. Follow-up that does not die. Manual follow-up usually stops after the first unanswered message. An agent re-engages quiet leads on a timed sequence and stands down the moment a person replies. It also flags warm leads going cold to the right manager, with a suggested next touch. Fewer no-shows. A booked meeting is only worth something if the buyer arrives. The agent confirms the slot, reminds the buyer 24 hours and 2 hours before, and re-books anyone who misses it. Closers who actually close. When first questions, budget checks and scheduling stay with the agent, senior salespeople spend the day with buyers who are ready. Spaceplace reports that around 60% of the top of the funnel is handled without a human. Before and after First reply. Without intelligent automation: hours or days, depending on who is on shift. With AI agents: seconds, 24/7, in the buyer’s language. Qualification. Without: done by closers, recorded inconsistently. With: one question at a time, scored, written to the CRM. Channels. Without: threads scattered across personal messengers. With: one deduplicated record per buyer. Follow-up. Without: stops after the first message. With: a timed sequence until the buyer replies or opts out. Meetings. Without: no-shows discovered on the day. With: confirmed, reminded at T-24h and T-2h, re-booked. Handoff. Without: the closer starts from scratch. With: the closer gets a briefed, scheduled buyer. One lead, from midnight to meeting A buyer messages your Instagram at 23:40. The agent replies within seconds, in the buyer’s language, and by 23:44 it knows the budget, the timeline and what they are looking for. At 23:46 a clean CRM record exists and a call is booked for the next day. At 9:00 your manager opens the CRM to a qualified, briefed buyer. Nobody on the team was awake for any of it. What automation should not decide Intelligent automation does not replace your team’s judgment. Pricing exceptions, sensitive accounts and large deals should reach a person, with the thread and history included. In Spaceplace, these thresholds are set by you and enforced technically rather than by a prompt, and every action lands in an audit trail: who or what acted, when, and on which rule. How to start Do not automate everything at once. Pick the most painful workflow, usually first response and qualification, and measure it before and after: First response time, by channel and by hour Share of inbound leads captured in the CRM Lead-to-meeting or lead-to-viewing rate No-show rate Hours your closers spend on unqualified conversations Test the agent against your own scenarios before it goes live, and again after every change. Once the numbers hold, extend the same agents and the same memory to follow-up, documents and after-sale service. See it on your own pipeline. Book a 30-minute live demo with your domain, your systems and one real workflow.

    • Use Cases
    • Industry Insights
  • 1 min read

    What Is an AI Agent and How Does It Work?

    Artificial intelligence is no longer just about answering questions. A new generation of AI systems — called agents — can plan, decide, and act on their own to complete complex tasks. But what exactly is an AI agent, and how is it different from the AI tools you may already be using? The basic idea An AI agent is a software system that perceives its environment, processes information, and takes actions to achieve a specific goal, often without a human having to guide every step. Unlike a chatbot that responds to a single prompt, an agent can break a task into steps, execute each one, evaluate the result, and adjust its approach along the way. Why it matters now Three things changed at once: models got good enough to plan, tools became callable from a conversation, and the systems around them learned to keep a record. Together they turn a clever answer into finished work.

    • Use Cases
    • Product Updates