Last Week in ConTech - 15 June 2026
How enterprise software interfaces may become customizable
Deep Insight: How enterprise software interfaces may become customizable
What if software adapted to each user, instead of forcing each user to adapt to it?
A growing opportunity with AI is the ability to build enterprise software that can be configured at the level of the individual user, project, or business. Not just through settings and permissions, but through workflows, interfaces, and operating logic that are customized to how each organisation or project team actually works.
For example, tools like Procore or Aconex define structured ways of working and enforce specific process flows. While the standardisation improves consistency, it often does not reflect the variations in how teams operate creating friction in adoption.
Additionally, when users request even small changes, such as adjusting a workflow or modifying a UI element, they are often asked to submit a support ticket.
The reality is that the request never gets actioned as it’s too niche or specific to one company. Enterprise software is built for scale across many customers.
It means companies are forced to adapt around tools, building workarounds and parallel systems. It’s also one reason why Excel remains so dominant. The flexibility allows teams to model how they actually work rather than how software expects them to work.
This week, Dalton Mills, a US startup, emerged out of stealth aiming to address this gap.
They are building a platform which allows trades companies to build custom operating systems using natural language, describing themselves as “Replit for the trades, where the operator builds the software their business runs on.”
It means teams can describe how they want their software to work and have those changes implemented immediately. As they describe it:
A roofer…opens Dalton and types one sentence…when the pitch is steeper than 9/12, add forty percent to the labor line, and flag anything over 10/12 for a safety review.
Or
Remember that decline button your old software would not let you put on an invoice? You ask for it in plain words, and it is there instantly. A change you would once have filed a ticket for and never seen, now takes about as long as it took to describe.
More broadly, it reflects a new opportunity created by AI: the ability to build bespoke software at scale.
Historically, custom or project specific software was expensive because every workflow, interface, and business rule had to be manually designed and coded. AI lowers that cost by allowing software to be generated and modified directly from natural language.
At a point-solution level, this is easier to understand.
Take an AI system that automates inspection report generation. It can be broken into three layers:
Data model (what information is required)
For an inspection report, this includes structured fields such as location, asset inspected, findings, severity, photos, corrective actions, and responsible parties.Workflow logic (what happens with that information)
Once data is captured, the system generates the report, applies templates and standards, flags issues for review, and routes actions to the appropriate stakeholders.Interface layer (how users provide the information)
This defines how information is captured, displayed, and consumed by users.
Traditionally, the last layer has been implemented through forms, dashboards, and mobile apps for data capture and display. In an AI-native model, it becomes flexible and multi-modal and designed by the user.
Because the system already knows what data is required, users can choose how they provide it. That input could come through WhatsApp, a voice agent, email with photos and notes, a web interface, or a structured form.
The shift here is that the consumption surface is configurable.
The backend remains structured and deterministic, while the interface becomes adaptive. It means that the data is decoupled from the interface entirely, with flexibility in both how it is captured and how it is displayed or consumed based on user preference.
The same principle applies to the workflow layer.
In traditional software, workflows are often predefined and difficult to modify. In an AI-native model, they can become configurable, with custom business logic applied on a per-job, project, or company basis. Different standards, review requirements, approval paths, and follow-up actions can all be implemented while using the same underlying data model.
I’ll be interested to see how this approach evolves beyond highly structured use cases and into more complex workflows that are collaborative, ambiguous, or exception-driven.
But given the direction of the market, the future of adaptive software that meets users where they are seems closer than we think.
In this issue there are:
10 Startup Fundings
2 Startup Emerged from Stealth
13 Policy and Regulatory Changes
9 New National Infrastructure Projects & Priorities
0 New investment funds
0 Acquisitions
6 News articles
64 new jobs posted - view here
Reading time: 14 mins
Startup Funding
AI
Dalton Mills, a US startup, raised $9.2m in funding. They are building an AI operating system for home service businesses across trades such as plumbing, electrical, HVAC, and more which allows operators to create and customize software using their own language and industry expertise, enabling workflows that reflect how their business actually operates rather than forcing teams into rigid, one-size-fits-all systems. More here.
Notes:
Dalton Mills is building a community and expert based go to market motion.
They have a private group called ‘Blue Collar Builders’ with trades owners who share notes on how they run their businesses, how they hire, how they price and what’s working.
As part of this they also run in person networking events and share writing, interviews and benchmarks on how homes services business actually run.
I like this approach as media / community can be an incredibly valuable way of building an engaged audience and distribution.
The challenge is that these community businesses often struggle with monetization.
Software, on the other hand, has scalable business models but faces rising customer acquisition costs.
This makes them strong partners and is a reason why large companies like HubSpot are acquiring media outlets.
Inspection
ConeLabs, a Canadian startup, raised $1.5m in Pre-Seed funding. They are building an infrastructure inspection platform that converts field imagery into inspection grade 3D models, automatically identifying defects and generating reports. More here.
Risk
Enlaye, a Boston startup, raised $5m in Seed funding. They are building an AI native risk lifecycle management platform which enables builders, owners, insurers and developers to identify risks early and mitigate their impacts across all stages of the project lifecycle. More here.
Finance
Earlytrade, an Australian startup, raised ~$10m in funding with Brick & Mortar Ventures investing. They are building an early payment platform which allows contractors to put idle working capital to work while providing subcontractors with earlier access to payments. More here.
Notes:
Under this model suppliers agree to receive slightly less than the full invoice in exchange for earlier payment.
The buyer determines which suppliers and invoices are eligible for early payment.
Suppliers then review approved invoices and estimated payment dates in their portal and decide whether to offer a discount and at what rate to be paid early.
These discount offers are evaluated against the buyer’s custom return goals and the dynamic discounting market.
The buyer then pays suppliers through their existing payment process, directly from their own accounts.
Compliance
Structured AI, a New York startup, raised $4.2m in Seed funding., They are building an AI platform which automates quality assurance and quality control processes, having developed an optical recognition solution which can complete compliance checks across an entire drawing set as well as provide resolution recommendations. More here.
Green Materials
Foamlab, a Dutch startup, raised €3m in funding. They are building high performance foams made from bacterial cellulose that are lightweight, durable and compostable and can be used for a variety of applications including insulation. More here.
Workforce Development
Skillmaker.ai, North Carolina startup, raised $4.5m in Seed funding. They are building an AI powered training and productivity platform for the skilled trades which combines automated knowledge collection with extended reality to deliver immersive, hands-on training experiences as well as using VR headsets and smart glasses to allow users to watch training material while building, fixing or updating the physical hardware in front of them. More here.
Notes:
They have initially gained traction in the automotive industry.
Data centers, telecommunications, HVAC, and home services are listed as their next focus areas.
Fleet Management
CameraMatics, a Dublin startup, raised $52.9m in funding. They are building a video monitoring and driver alert system for commercial fleet operators across industries including construction that improves safety and manages compliance including providing digital equipment checks, assists with asset utilization and preventive maintenance, automates compliance and monitors fuel usage and more. More here.
Other - EV Charging
InCharge Energy, a Los Angeles startup, raised $46m in funding. They are a lifecycle infrastructure company, delivering EV charging, electrical and lighting and distributed energy solutions supporting customers from initial concept and construction through long term ongoing operations and maintenance by acting as a system of record for warranties, service contractors and preventative maintenance programs. More here.
Notes:
I’ve been noticing an increasing number of solutions attempting to bridge the gap between construction and asset operations.
One reason is that valuable project information often becomes fragmented or difficult to access during the asset handover phase.
On a typical project, thousands of documents are created, including drawings, schedules, submittals, RFIs, inspection reports, commissioning records, closeout documentation, and more.
These documents contain information on materials, installed systems, performance criteria, warranties, maintenance requirements, and compliance status.
At handover, this is often delivered to the owner’s operations team as PDFs, spreadsheets, and other files.
It can be incredibly difficult to sift through this information and locate the specific details needed years later when a warranty claim, maintenance activity, or asset issue arises.
Knowledge can also be lost as project teams change roles or leave organisations, making it harder to know where critical information is stored.
As a result, startups are increasingly recognising that if they can structure and organise this information during the construction phase, they can facilitate the handover and become the system of record during the operations phase.
This allows asset information to remain connected, searchable, and accessible throughout the life of the asset.
The information can then be queried for downstream activities such as maintenance planning, warranty management, compliance tracking, and asset servicing.
For software providers, this creates the opportunity for long-term customer relationships that can extend well beyond project completion and, in some cases, for the operational life of the asset.
Other - Marketing
Artis, an Austin company, raised $7.3m in Seed funding. They handle marketing and brand strategy running the social media, websites, and AI search presence for residential architects, interior designers, and custom home builders and have developed a technology platform that is expanding from social media into websites and generative engine optimization. It is fed by the same underlying intelligence built for each specific practice being its voice, its standards, the decisions and craft behind every project. More here.
Out of Stealth
Piper AI, a San Francisco startup, emerged out of stealth. They are building an end to end platform for construction pursuit and preconstruction teams, streamlining the bid management process by ingesting bid packages, analyzing to surface risks, gaps and insights and then assisting with developing a clear strategy on what to bid, how to price and why. View the website here.
American Terraforming Corporation, a US startup emerged from stealth. They are building next generation heavy equipment designed for an era of physical AI. View the website here.
Policy and Regulatory Changes
US
Republicans probe China’s influence in data center opposition
House Energy and Commerce Republicans are claiming there is “strong evidence” that growing opposition to data centers is due, at least in part, to foreign influence campaigns linked to China.
Republicans asked the President’s Council of Advisors on Science and Technology and FBI Director Kash Patel for a briefing by June 18.
This is to help lawmakers “better understand how the Administration is investigating the foreign influence campaigns and taking action to mitigate these efforts.”
Voters in California City Favor Permanent Ban on Data Centers
Voters in Monterey Park, Calif., appear to have overwhelmingly approved a ban on data centers.
The unofficial total has 86% of the vote in favor of the ban.
This would become what is believed to be the first city in the nation where residents passed a measure permanently prohibiting.
Americans Now Overwhelmingly Oppose New Data Centers Near Them
When asked last September how people would feel about a nearby data center project, 55% of Americans “strongly” oppose a data center project built near where they live, and an additional 16% are “somewhat” opposed.
Only 21% of Americans would support a new nearby data center.
House passes bill to speed up union-employer negotiations
The House of Representatives passed the Faster Labor Contracts Act.
The bill would accelerate negotiations between employers and workers who have voted to form or join a union.
When workers vote for labor representation, the resulting negotiation can drag out.
The bill cites a 2021 Bloomberg Law study finding it took 465 days on average between the vote and final contract approval.
The result of the bill, should it become law, would create deadlines for specific negotiations and recourse for workers to settle disputes, tipping scales in favor of unions and other labor organizers.
Virginia General Contractors to Face Expanded Responsibility for Unpaid Wages
Beginning July 1, 2026, potential liability of general contractors in Virginia for unpaid wages will greatly expand under newly enacted House Bill 238.
Historically, Virginia law already imposed some degree of joint liability on general contractors for subcontractor wage violations.
Since 2020, GCs could be held jointly and severally liable for unpaid wages owed to subcontractor employees, but only if they “knew or should have known” about the wage violations.
This knowledge requirement served as an important limitation, allowing contractors to rely on contractual protection, certifications, and reasonable oversight practices.
The new legislation largely removes those limitations.
In New Jersey, Sherrill Agrees to Delay Protections Against Sea Level Rise
The New Jersey Governor has agreed to a one-year delay and new review of regulations intended to prepare the state for sea-level rise and more flooding.
The regulations, known as the Resilient Environments and Landscape (REAL) rule, factor in an ever-increasing rise in sea level by 2100.
They mandate more stringent construction standards and flood protections, better stormwater management and greater protections for wetlands.
The most controversial provision requires new shore homes and substantial remodels of existing homes to be constructed four feet higher than current FEMA standards.
Texas governor talks tough on data centers, calls for clampdown
The Texas Governor released a broad set of proposals aimed at regulating the growth of data centers.
He said in a letter to state electricity officials that the unprecedented growth of data centers and their major energy needs require “oversight” to shield Texans from higher electricity costs.
State electricity officials have been directed to send him a memo on actions and recommendations by July 17.
The governor also said he would work with state lawmakers during next year’s legislative session to pass laws that require data centers to pay for their own electric infrastructure and use water-efficient technologies.
He also called for repealing sales tax exemptions “and other outdated or unnecessary incentives for data centers.”
Texas county rescinds its data center moratorium after $100 million lawsuit from developer
A rural North Texas county that appears to be the first in the state to pass a data center moratorium has rescinded the measure and adopted a checklist it will require of data center developers.
This comes after being sued by a developer for $100 million in damages.
New Arizona law takes aim at unnecessary permit delays for home construction
Under Senate Bill 1566, municipalities and counties must have set time frames for issuing licenses or building permits for home construction.
If municipalities or counties are found to be intentionally obstructing the approval of licenses or creating delays out of malice, they will face a $5,000 penalty.
Connecticut Governor signs the state’s omnibus 2026 solar bill
The Connecticut Governor has signed HB 5340, a sweeping energy policy bill.
It extends the state’s long-running Renewable Energy Solutions programs through 2035.
It also launches a successor community solar program, requires automated residential solar permitting and establishes new rules to allow for residential use of portable, plug-in solar panels.
The bill also immediately places a moratorium on certain large-scale solar development and will limit the total annual funding and capacity available to new solar installations in the state.
FERC approves PJM fast-track review for ‘shovel-ready’ power projects
The Federal Energy Regulatory Commission on Tuesday approved the PJM Interconnection’s proposal for an “expedited interconnection track” for large generating projects.
Under the process, PJM will consider up to 10 interconnection requests a year on a fast-track basis.
This is for new or uprated capacity resources of at least 250 MW that can come online within three years.
The process is set to expire at the end of 2027.
Could Florida Really Eliminate Property Taxes?
The property tax plan for the November ballot would raise the homestead exemption to $250,000 and require the Legislature to enact a plan to eliminate property taxes entirely.
This would apply to the vast majority of Floridians who own the homes they live in.
Local governments are likely to lose tens of billions of dollars in revenue.
Local officials have warned that residents could face cuts to schools, fire and other services and even possibly charges to use public parks.
China
China summons state construction giants over safety failures
China’s national safety regulators summoned state-owned construction giants China Communications Construction Company and Power Construction Corporation of China.
It was what they said were safety failures linked to two major incidents.
The emergency management ministry stated that negligence, falsifications, illegal subcontracting, cutting corners and safety management failures at the two companies and their subsidiaries contributed.
National Infrastructure Projects & Priorities
US
Trump administration announces $850M to modernize US coal capacity, build 2 new plants
The U.S. Department of Energy committed up to $850 million to construct two new coal-fired power plants, modernize more than a dozen other facilities, and bolster coal export capabilities.
The funds will support 17 total projects, including new coal-fired plants in Anchorage, Alaska, and Mt. Storm, West Virginia, which will have a combined capacity of 2.85 GW.
Meta earmarks $115M for workforce academy to support data center construction
Meta has launched America’s Workforce Academy, an initiative designed to provide free skilled-trades training and guaranteed job opportunities on data center construction projects.
The program, which Meta called the largest private-sector commitment to skilled-trades training with a job guarantee in U.S. history, will launch in Indiana, Louisiana, Ohio and Texas, this year.
Google vows $50M for skilled trades training
The money will support 14 labor unions and four trade and contractor associations, aiming to reach 300,000 workers across 20 states.
The goal is to help level up training and outreach, which will better educate a new wave of workers to help meet the sky high demand on construction projects, including data center builds.
China
China Preps $295 Billion Plan to Fund Nationwide AI Buildout
China is preparing to spend around 2 trillion yuan ($295 billion) over the next five years on building data centers.
Key government agencies including the National Development and Reform Commission are drafting a blueprint to erect a network of inter-connected computing hubs across the country.
State firms such as China Mobile Ltd. and China Telecom Corp. will operate the bulk of the data centers and ensure they’re connected.
China Starts Prefabricated Power Hub for Data Centers
China’s first prefabricated computing power hub has started operations, offering a faster and lower-cost way to build and supply electricity to data centers.
The system can cut land use by more than 30% and reduce overall costs by about 20%.
Construction can be completed in as little as five months.
China’s nuclear power capacity nearly doubled since 2016
From 2016 to 2024, China’s nuclear generation capacity increased 76% (24 GW).
China added an additional 1.1 GW of nuclear power capacity in 2025 and 2.2 GW in 2026 (through May).
They have 36 reactors under construction, accounting for more than 49% of total world nuclear construction.
Japan
Japan Aims To Replace 14 Ageing Nuclear Reactors By 2050s
Japan plans to rebuild between two and five ageing nuclear reactors by the 2040s and as many as 11 to 14 by the 2050s.
The proposal reflects a shift towards greater reliance on nuclear energy to help meet rising power demand and reduce costly fuel imports.
It is the first official numerical target for reactor replacement released since the March 2011 accident at the Fukushima-Daiichi nuclear power.
India
AirTrunk commits $30B to build 5GW of AI data centers in India
AirTrunk said it would invest $30 billion in India by 2030.
Data center capacity in India is projected to rise to as much as 8GW by 2030 from ~1.5GW today.
Earlier this year, New Delhi offered foreign cloud providers tax exemptions through 2047 on services sold overseas if those workloads are run from Indian data centers.
Thailand
Thailand’s Richest Man Plans $4.3 Billion Expansion Amid AI Boom
They aim to add as much as 2,000 megawatts of data center capacity over the next 5 years to meet rising demand for AI and cloud-computing services.
News
Meta steals a tactic from Tesla and builds data centers in tents
Inside the AI Boom’s Arctic Outpost
Washington’s Best Housing Idea Has an Image Problem
Inside the biggest bet in corporate history (Bessemer Venture Partners)
By next year, capex from the big five (AMZN, MSFT, GOOGL, META, ORCL) are expected to cross $1 trillion per year.
Cumulatively from 2025 to 2031, roughly $8 trillion will go into the ground
The Physical Seams of the AI Buildout
The Mag Seven’s 2026 capex guidance is larger than the Apollo moon program, the transcontinental railroad, and the interstate highway system combined.
Tall buildings lead to more compact and productive cities
This article analyses over 11,000 urban agglomerations and 300,000 tall buildings to explore the effect of height restrictions on welfare.
If I missed anything this week, please reply and let me know! I’ll make sure to include it next week.