Startup Intelligence

A look at how startups today are building the intelligence infrastructure layer of the economy.

If you want to get a sense of the future, emerging companies offer strong signals. They’re particularly useful for trend spotting because they are often run by people who are anticipating change before the broader market has.

We fill the PSFK graph in different ways, some human, some algorithmic. Our startup research leans towards the latter - with search algorithms scanning for new companies to add to our trends database. What’s useful about the machine approach is that it doesn’t get inhibited by language, region or topic like a human does - it just gets to work like a little Sherlock.

So… by running pattern recognition across the data “we” collected over the last 120 days, we identified 52 different trends for start ups. The broader macro themes you’ll find in this analysis include:

• Monitoring & Diagnosis - decision making systems built on computer vision and sensor data

• DropIn Inputs - Manufacturers make good products with good ingredients

• End to End Operating Systems - workflows that run across the business lifecycle

• Trust & Proof - infrastructure around auditing and compliance.

• eShoring - the offloading of compute and data

• Moment Commerce - anticipating intent and accelerating sales

• Simulators - Simulation becomes the prime R&D approach

But what’s the broader pattern here? Where is business going? What do all these trends tell us. I’d love to hear your feedback - but here’s my take:

Start ups are clustering around intelligence. The are working on ideas that:

1. Feed intelligence
   More sensing, more signal, more measurable inputs into systems.

2. Sustain intelligence
   Infrastructure that keeps intelligence running reliably, securely, and at scale.

3. Trust intelligence
   Work that is auditable, provable, and regulator-ready. Intelligence that cannot be trusted cannot scale.

4. Use intelligence
   Intelligence converted into action, decisions, and transactions.

Interestingly, the startup landscape isn’t clustering around industries. It’s clustering around the lifecycle of intelligence itself. Maybe…. we could say that start ups today are building the intelligence infrastructure layer of the economy.

Piers, Founder, PSFK

(For more : check the PSFK newsletter on wearables and a planetary intelligence system)

Key Trends In Start Ups

Trust & Proof

One group of startups are building the infrastructure around auditing and compliance. Their systems make claims, transactions, and AI outputs trackable with evidence, logs, provenance, and regulator-ready reporting baked in. The shift is from “best-effort reporting and opaque automation” to audit-ready systems-by-default, because enterprises can’t scale AI, climate disclosures, or new payment rails without provability.

Examples and Trends within this theme of trust and proof include:

• AIRMO within the trend of Template-Driven Compliance Reporting Stacks With Audit Trails

• Bridge within the trend of Programmable Stablecoin Rails for Business Payments

• Floodlight - Audit-Grade Carbon Accounting Pipelines

• Divide Graph - Evidence‑Linked AI Outputs for Compliance and Assurance

• Definely - Legal Workflows Embedded Into Primary Work Surfaces

• Zenity - AI‑agent governance and GRC systems for enterprise agents

• ClimateCamp - Audit‑grade climate measurement and reporting stacks

• Tempo - Compliance‑first crypto infrastructure stacks

Moment Commerce

When it comes to commerce, we also see young companies exploring Moment Commerce. Startups are collapsing the distance between attention/intent and purchase by embedding automation directly into the channels where intent is expressed - whether that’s in social, messaging, discovery flows, and content moments. The shift is from optimizing clicks to optimizing completed transactions, often with payments, loyalty, and upsells integrated into the same flow.

Examples and Trends within this theme include:

• Syncpop within the trend of AI social-to-booking autopilots for local SMBs

• ExpectMe - Booking-to‑Upsell Monetization Layers for Hospitality and Travel

• Diddo - Content monetization that turns moments into transactions

• Buzz - Travel booking that merges discovery, service and payments into one flow

End to End Operating Systems

Within this theme we see startups building what are, in effect, Vertical AI Operating Systems. New companies are bundling fragmented tools into single operating surfaces that run a business end-to-end - workflow execution, data capture, billing, reporting, and increasingly payments/finance. Instead of “add another point solution,” buyers get a system-of-record that automates reconciliation and becomes the platform for AI, embedded finance, and new services.

Examples and Trends within this include:

• Viz within the trend Hospital AI Workflow Layers Replacing Single‑Point Tools

• Harbor Lynx - End-to-End Operating Systems for Niche Physical Operations

• Nominal - Multimodal Ops Observability for Physical Assets Replacing Spreadsheet Maintenance

• HerStage - Women’s Health Platforms Bundling Lifecycle Care With Virtual Coverage and Coaching

• Brigit - Nonprofit RevOps Automation

• Woliz - Vertical Operating Systems for Field‑Heavy Industries

• Clovo - Condition‑Specific Digital Therapeutics With Always‑On AI Coaches

• Cell - Chat-native operational workflows that replace portals for routine tasks

• KaHero - SME operating systems bundling POS, payments and embedded finance

eShoring

It’s probably not the best name but I was trying to describe how startups are clustering around offering off-premise data and computing operations fueled by a collision of AI workload demands with constraints around cost predictability, latency, and sovereignty. My AI wants to call this Sovereign Compute & Productized Deep-Tech Infrastructure: So we see startups turning AI infrastructure from a bespoke engineering project into a purchasable product: managed performance, packaged inference stacks, and deploy-anywhere configurations (on-prem, offline, air-gapped).

Examples and Trends within this theme of eShoring include:

• Coreweave that sits within the trend of AI-Native Compute and Private Model Deployment Replacing General Cloud Setups

• Groq - Specialized AI Compute Clouds Sold as Managed Performance

• Cohere - On‑Prem and Offline AI for Sovereignty and Denied Networks

• QuEra - Quantum and Photonics Components Moving Into Procurement Catalogs

• Sarvam - On‑prem and offline‑first deep tech for restricted or sovereign environments

• TensorWave - Packaged AI compute and inference as a product (cloud + hardware bundles)

Monitoring & Diagnosis

In the data I saw how startups exploring “validated decision support” and embedded care. These entrepreneurial companies are shipping narrow, validated diagnostics and monitoring modules that work inside clinician workflows (viewers, EHR pathways, intraop systems) and extend into the home with continuous interpretation. The pattern is moving from standalone analytics to in-situ decision support with clear accountability and adoption pathways.

Examples and Trends within this include

• Rivana - AI-Guided Procedural Imaging Making Specialist Decisions Repeatable

• LEO - Consumer Self‑Verification Biomarkers Moving Into Everyday Routines

• Qure - Clinician‑Facing AI Diagnostics and Monitoring Modules

• RingConn - At‑Home Diagnostic Hardware as Subscription Services

• CergenX - Breakthrough-designation imaging that makes staging and perfusion visible during procedures

Drop-In Inputs for Sustainable Manufacturing

I guess, if you want to make better-for-the-planet products, you better look at the stream of ingredients and energy that feeds your production line. In this theme of Decarbonized Inputs That Fit Existing Supply Chains, startups are scaling new materials and production methods by licensing into existing supply chains (instead of trying to replace them). The emphasis is on drop-in adoption: converters, machinery/process licensing, and distributed manufacturing approaches that reduce carbon and dependency on centralized petrochemical infrastructure.

• Solidec that fits within the trend of Electrified Modular Reactors Replacing Bulk Chemical Supply

• Lite1 - Bio‑Grown Industrial Inputs Replacing Petrochemical and Mined Feedstocks

• Pulpac - Plastic‑Free Packaging Tech Licensed Into Existing Supply Chains

• Pulpex - Fiber and biomaterial packaging shipped as a licensed manufacturing process

Simulators

I also see a pattern of start ups exploring simulation-first R&D & digital twin tech. Startups are shifting simulation to the primary research approach: in biology (which experiments to run, which candidates to advance) and in complex environments like the OR (training twins, workflow standardization). The shift is from “wet lab / real-world first” to compute-first prioritization that compresses iteration cycles and standardizes execution.

• Caresyntax that fits in the trend of OR Digitization With Intraoperative Guidance and Training Twins

• Turbine - In‑Silico Bio Platforms Acting as Lab Capacity

• Owkin - Biology R&D Shifts to Simulation and Automation Platforms

• Clovo - Condition‑specific digital therapeutics with built‑in validation loops

• Insilico - Computation-first biology platforms that prioritize simulation and multimodal models

***

As I said earlier, Start ups are clustering around intelligence. The are working on ideas that Feed, Sustain, Trust and Use intelligence

Interestingly, the startup landscape isn’t clustering around industries. It’s clustering around the lifecycle of intelligence itself. We could say that start ups today are building the intelligence infrastructure layer of the economy.

If you’d like to deep dive into an element of this research or receive a presentation, contact PSFK.