Until recently, portals sat at the centre of the property transaction ecosystem because they aggregated inventory, consumer attention and leads. But a growing stack of software, AI-native discovery tools and transaction platforms is giving agents, sellers and buyers alternative ways to generate demand, discover properties and complete transactions without relying as heavily on portals.
We mapped > 80 tools and platforms across the real estate technology landscape helping to make this shift possible. What did we learn?
Agents are becoming more capable without the portal
Historically an agent won an instruction, created the listing, uploaded it to the major portals and waited for enquiries to arrive. The portal provided the distribution; the agent managed everything that happened after the lead.
Software now offers real estate agents tools and services to be more proactive, with implications for real estate agents and portals alike.
CRM, chatbots and automated outreach
Modern CRMs and chatbots are helping agents automate the customer relationship and activate their existing databases. An agent who can use AI to identify a likely seller off portal, and nurture that relationship has lessened its reliance on the portal.
Brokerages already sit on valuable assets: years of buyer, seller, property and interaction data. Historically, much of that data sat passively inside a basic CRM system. AI-first CRMs continually analyse and activate the same data as relevant listings are mandated thus making it much more useful for the agent.
Platforms such as Structurely, Street.co.uk, Coraly, Koqi, and Lofty are doing exactly this.
On the sell side, Lofty can analyse an agency's existing CRM to identify customers most likely to transact, automate outreach across channels and recommend next actions. Other tools respond to enquiries, qualify buyers by budget, timing and financing readiness, match them against relevant properties and automatically follow up around viewings or potential offers.
The CRM therefore starts to evolve from a system of record into a source of demand while also serving as a tireless assistant.
Creating and marketing inventory is getting cheaper
Listing creation tools have drastically reduced the cost and effort of marketing property inventory.
Companies such as Collov AI, Pedra, ApplyDesign, RealEstateContent.ai, ReimagineHome and Virtual Staging AI can help with listing content, staging, imagery and other marketing assets at a fraction of what a photographer and a copywriter used to cost.
Generating leads beyond the portal
These tools are trying to generate the buyer or seller opportunity itself. If technology can identify or generate those consumers elsewhere, then the portal has more competition for the agent's acquisition budget.
Seller lead generation
Platforms such as Offrs, Casafari and DeepImmo help agents identify homeowners who may be preparing to sell. These signals can come from AVM interactions, website engagement patterns, listings with price reductions or having been a long time on the market, or an agency's CRM and third-party datasets. then gives the agent the available contact details to start acquisition outreach.
Unlisted starts with generating buyer demand for off-market properties. Buyers can express interest in homes that are not currently for sale and build a waitlist around a property. That latent demand can then be surfaced to the homeowner, potentially creating a seller opportunity before they had planned to go to market.
Buyer lead generation
Platforms such as Jitty, Omny and Holofy are rebuilding property search around natural-language intent rather than traditional portal filters.
Instead of asking buyers to choose location, price, bedrooms and property type, these products can increasingly understand questions such as:
Find me a three-bedroom home with good natural light, within 30 minutes of my office, close to a strong primary school and in an area that should be easy to resell in five years.
Buyers can essentially search for qualitative property characteristics, neighbourhood attributes, commuting requirements and other factors that are difficult to express through conventional filters.
The commercial model can also look very different from a traditional portal. Jitty, for example, removes the traditional listing subscription barrier for agents, with properties listed for free and monetisation pushed toward downstream services. While Omny connects conversational buyer intent with the live inventory of professional agents and aims to deliver fewer, more qualified enquiries rather than a high volume of low-context leads.
AI-native property discovery
If consumers increasingly begin their property journey inside ChatGPT, Gemini, Claude or Google AI mode, agents get an alternate channel to capture leads beyond portals
Companies working on GEO, conversational search and AI-native discovery, including Kleio, Qvery, Propelos and Tely.ai enable agencies to get discovered in the interfaces.
Propelos, embeds natural-language property search into agency websites and optimizes the agency website for GEO. Qvery tracks whether an agency or its listings appear for relevant AI searches and maps the sources AI engines cite. Tely.ai creates hyper-local content intended to generate traffic from both conventional and AI search. Kleío goes further by turning agency databases into AI-readable knowledge layers and connecting structured, live inventory into LLM interfaces.
The most ambitious companies want to own the transaction
AI-native transaction platforms
The previous generation of tech-enabled brokerages, likes of Casavo, McMakler, Purplebricks and Compass among them, promised to digitise real estate agencies. They moved processes online, reduced branch overheads, introduced flat-fee packages and built better agent software.
However, the economics often remained difficult, customer acquisition was expensive, transactions remained people-heavy, and the service quality was difficult to standardise across individual agents.
AI potentially changes that equation. An AI-native brokerage can automate materially more of the work itself. Platforms such as Turbohome and Homa combine AI-native search, property analysis and transaction coordination with human agents used at the points where they add the most value.
Turbohome, backed by FJ Labs and Sequoia Capital, for example, combines a conversational buyer copilot with an economic model that returns part of the traditional buyer-agent commission to the consumer.
AI-native transaction marketplaces
Rather than building its own large agent workforce, a platform can provide the software, matching and transaction layer while relying on an existing distributed network of independent agents.
Zefir is an early example, it pools listings and buyer demand across independent agents, while its AI-powered buyer assistant helps consumers search, compare and decide. The transaction itself is then executed through agents in the network.
This gives an AI-native transaction platform the benefits of local human execution without recreating the entire fixed-cost structure of a traditional brokerage.
FSBO
FSBO 1.0 largely digitised distribution. A homeowner could place a sign outside the property or put the home online, but still had to price the property, prepare the listing, manage enquiries and viewings, negotiate offers, prepare paperwork and coordinate the closing
The seller saved commission by doing the agent's work themselves. An AI-native FSBO platform can potentially help price the home, create the listing, answer questions, coordinate viewings, compare offers and guide the seller through the process, bringing in licensed professionals only when necessary.
Ridley, founded 2025, replaces a percentage-based seller commission with a transparent flat fee, while an AI copilot guides the seller from pricing through offers and closing.
What does the next decade look like for real estate agents?
AI is unlikely to eliminate all real estate agents, but it could dramatically change how many agents the industry needs, which agents win, and what an agency looks like. If a top agent equipped with AI can respond instantly, understand every lead, maintain thousands of relationships, produce better marketing and manage more concurrent transactions, there is progressively less room for an agent completing a handful of transactions a year with limited differentiation.
Over the next decade, transaction volume is likely to consolidate around a smaller number of highly productive agents who use AI across prospecting, customer management, search, marketing and transaction execution.
In our view, the strongest incumbent agents will adopt these tools quickly while part-time and low-productivity agents will find it harder to compete. A new generation of tech-enabled brokerages may use AI to automate enough of the transaction to finally deliver the economics that TEB 1.0 promised.
