The Title You Have to Outgrow

Myriad360

The asset in a systems integrator never appears on a vendor price sheet. It is the client calling before the architecture is decided, while AI, security, cloud and budget are arguing in the same room.

One Equity Partners has acquired a 51% stake in Myriad360, giving the New York-based technology integrator more capital for organic growth, talent, services and acquisitions. The parties did not disclose the transaction value. Myriad360's leadership will remain in place, and management will retain a significant ownership position.

That ownership detail matters. Andrew Fisher founded Myriad360 in 2003 and now serves as Founder and Chief Culture Officer. Jay Miley remains President and CEO, while Herbert Hogue remains CTO. They are betting that a 23-year-old operating culture can carry more weight.

Myriad360 has already been testing that premise. It acquired Advizex and assets of F3 Technology Partners in 2026, widening its reach across AI-ready infrastructure, cloud, cybersecurity, managed services and healthcare technology. CRN estimates the combined business at $1.2B in annual revenue with about 570 technology professionals.

Enterprise technology is already complicated. AI plans depend on data architecture, cybersecurity, cloud economics, networking and implementation capacity behaving like one system even when the vendors do not.

That is where the channel earns its keep. A good integrator does more than resell equipment or collect partner badges. It decides which pieces should work together, owns the ugly handoffs and stays close enough to notice when the strategy and operating environment have stopped matching.

OEP knows that playbook from prior IT-services investments. Partners Charlie Cole and Carlo Padovano are bringing capital and acquisition experience. Myriad360 is bringing the relationships, engineering bench and vendor-agnostic position that can make those combinations valuable instead of merely larger.

Now the risk moves to execution. Miley says Myriad360 has 3 engineers for every salesperson. Acquisitions can deepen that bench, but they can also create overlapping processes and a client experience that feels assembled instead of integrated.

The next phase will be measured in moments no press release can stage: whether an acquired engineer can solve a legacy client's problem and whether new services reach the field without slowing it down.

Scale gives Myriad360 more doors to walk through. The company still has to make every room feel like the client called the right one.

Beatport

Professional DJs kept buying downloads while the rest of consumer music moved toward streaming because a track is not only something they hear. It is working inventory: purchased, organized, edited, loaded into performance software, and expected to behave when the room is full.

Square Nine Capital has now invested in Beatport. The financing refinances an existing credit facility and adds capital for the platform, products, and wider creator ecosystem. The amount, loan terms, and valuation were not disclosed, which matters because this is a capital-structure story, not a conventional venture round with a tidy number attached.

Square Nine specializes in private credit for middle-market companies. Its role gives Beatport more flexibility while Axar Capital remains part of the picture after years of rebuilding the business from the wreckage of SFX Entertainment.

Beatport says it supports more than 80,000 labels and 400,000 artists, has 10M registered users, and has paid $400M to artists and labels since 2004. The 2026 IMS report says revenue grew 9.4% in 2025, extending annual growth of 5-10% since 2017.

The download store is still central, but the company around it now reaches streaming, DJ software, label services, producer tools, media, discovery, and events. Beatport kept moving closer to the work of DJs, producers, artists, and labels instead of becoming another passive listening service.

Matt Gralen became CEO in July, Helen Sartory became President and COO, and Robb McDaniels moved to Executive Chairman after the post-bankruptcy turnaround. Co-founders Jonas Tempel, Eloy Lopez, and Bradley Roulier built the original store around a problem DJs could feel in 2004: digital performance was arriving before legal, reliable access to the files.

IMS and MIDiA Research value the global electronic-music industry at $15.1B in 2025, up 7%. Growth across recordings, publishing, software, events, and creator services makes the relationship between professional tools and culture more commercially important.

The financing leaves real questions. Beatport has not disclosed how much new money sits beyond the refinanced facility, which products receive it first, or the economics Square Nine underwrote. Those answers will arrive through execution.

Beatport's advantage is that its customer does more than press play. The next phase depends on how much more useful the company can become before, during, and after the moment a DJ chooses the track.

Command Alkon

Private equity can transfer control in a sentence. Concrete keeps hardening while the lawyers finish the paragraph.

Francisco Partners has completed its acquisition of Thoma Bravo's majority stake in Command Alkon. Heidelberg Materials remains a significant minority investor, keeping one of the world's largest heavy building materials producers inside the ownership group. The purchase price and other financial terms were not disclosed.

This is not software floating above an industry. Command Alkon sits inside the operating day of ready-mix, asphalt, aggregate, and cement producers. Command Cloud connects raw materials, production, dispatch, logistics, fleets, tickets, payments, and financial workflows. The company says more than 14,000 plants in over 80 countries rely on its technology.

That installed context is the asset. A late truck can push a pour. A poor batch can become material nobody wants. A disconnected ticket can keep a completed delivery from becoming cash. The software learns value by living close to those consequences.

Thoma Bravo acquired Command Alkon in 2020. HeidelbergCement, now Heidelberg Materials, took a 45% minority stake in 2021 while Thoma Bravo remained in control. Over that ownership period, Command Alkon moved toward Command Cloud, launched Batch AI and the Command Intelligence agentic platform, and expanded through acquisitions.

Francisco Partners is taking the baton after the architecture changed. Partners Petri Oksanen and Quentin Lathuille have pointed to infrastructure investment, agentic AI, and autonomous plant operations as the next opportunity.

Heidelberg Materials helps close that distance. Its continued ownership gives the partnership an industrial operator with a direct view into the plants and logistics Command Alkon serves. That is useful friction. It forces AI ambition to meet the schedule, safety, quality, and margin realities of physical operations.

Command Alkon President and CEO Martin Willoughby and CTO Charles Evans now carry a more demanding brief. The company has spent 50 years building industry knowledge and says its next chapter includes agentic systems, autonomous plant capabilities, global expansion, and payments. Each promise has to earn trust inside work that cannot pause for a software demonstration.

The ownership story is finished on paper. Inside a materials operation, the next chapter will arrive one decision at a time, while the trucks are already moving and the mix is already setting.

ElevenLabs

Private-company equity can become life-changing on a cap table while remaining useless at the closing table. The valuation rises, the number in the employee portal gets larger, and the person who helped create it still cannot spend a share that has no buyer.

ElevenLabs just built that buyer pool.

The company completed a $300M employee tender offer at a $22B valuation, double the $11B mark from its February Series D. Wellington Management and T. Rowe Price led the transaction, joined by institutional investors including Goldman Sachs, EQT, GIC, Andreessen Horowitz, ICONIQ and Lightspeed.

That $300M needs honest accounting. This was a secondary transaction, which means employees and other existing holders sold shares to investors. It creates liquidity for the people holding the equity; it does not automatically put $300M of new operating capital on ElevenLabs’ balance sheet.

The distinction makes the transaction more interesting, not less.

Mati Staniszewski and Piotr Dąbkowski founded ElevenLabs in 2022 around a voice problem they understood personally: dubbed media could carry the words and lose the person. The company began with text-to-speech, then expanded into ElevenCreative, ElevenAPI and ElevenAgents, which connects agents to business knowledge, systems and workflows.

ElevenLabs says enterprise customers now account for 55% of revenue. Its agents handle more than 15M conversations each week, up 3x since February, while ElevenAgents ARR has also grown more than 3x. Those company-reported figures show what the new investors are pricing: voice moving from a creative feature into business operations.

The employee side carries different evidence. ElevenLabs says the team has grown beyond 800 people across more than 50 countries. At that size, equity stops being an abstract recruiting promise and becomes a retention system that needs an exit valve before the company itself exits.

A tender lets some people turn concentrated private-company risk into real liquidity while the business keeps building. It also lets investors establish ownership without forcing another fundraising round.

The $22B valuation will attract the headline. The more revealing move is that ElevenLabs is trying to make staying for the next chapter compatible with realizing some value from the last one. The company is teaching AI to carry more human meaning through a conversation; its employees are deciding how much of their own work to keep carrying into the future.

Radical Ventures

A country can produce world-class AI companies and still rent the most expensive years of their growth from somebody else.

Canada helped build the research base, talent and founders behind companies such as Cohere, Waabi and Xanadu. When the next round required a different class of check, much of the late-stage capital came from the United States.

Radical Ventures is putting more Canadian institutional money into that handoff.

The Toronto-founded AI investment firm has secured well over $1B in commitments at the first close of the Radical Breakouts Fund. PSP Investments, CPP Investments, HOOPP, TD Bank Group, BMO Financial Group, CI Global Asset Management and OPTrust anchored the new strategy alongside other global investors.

First close matters. Radical has money committed and a strategy in motion, but it has not disclosed the exact total above $1B, a numeric final target, the final-close timetable or individual LP commitments. “Multi-billion-dollar” describes the ambition, not a disclosed total.

Jordan Jacobs, Co-Founder and Managing Partner, and Tomi Poutanen, Co-Founder and Partner, built Radical around specialist access to AI founders. Breakouts is about keeping that seat when the cap table gets expensive.

Radical says technology companies are staying private longer and creating more value before an IPO. The firm's first institutional fund dates to 2019, but later rounds can still force a specialist to hand ownership to larger pools of capital.

The Breakouts Fund is built for global AI scale-ups, not only Canadian companies. It is already investing in Discovery Loop, Etched and Crusoe. That makes the strategy less a patriotic holding pen and more a bid to let Canadian institutions participate wherever Radical believes the strongest AI companies are compounding.

The capital does not prove those companies become the trillion-dollar businesses in the pitch. It changes who can keep underwriting them while the outcome is still private, technically difficult and financially demanding.

For founders, that means a specialist investor may remain useful after the early round. For LPs, it means accepting late-stage AI exposure through a manager whose edge has to survive larger prices, more crowded rounds and less forgiving ownership math.

Canada's AI story has spent years celebrating what gets invented at home. Radical's first close moves the question downstream, toward who can keep buying into the companies once the science becomes a very expensive business.

Supabase

The database used to arrive after an application had earned one. An AI agent can create both before lunch, then repeat the trick a few million times.

A database starts behaving less like a machine attached to one application and more like an operating resource: cheap to create, isolated, disposable when the work is done, and capable of growing when a prototype becomes a business.

Supabase has raised $150M to build for that handoff. GIC led, with CapitalG, IronArc, and Square Peg participating. Supabase did not name the round or disclose a valuation. It said the capital will support agent-driven database development and employee liquidity.

The money arrived with a second decision. Supabase is acquiring Turso, whose team rebuilt SQLite in Rust and designed a cloud architecture that can load and suspend millions of small databases without dedicating a machine to each. Price and closing terms were not disclosed.

Co-founders Paul Copplestone, CEO, and Ant Wilson, CTO, have spent 6 years making Postgres easier to adopt without hiding the database. Turso gives them another infrastructure shape for work that arrives before Postgres needs to carry production scale.

Supabase reports more than 13M developers, plus 1M users and 4M databases added per month, with agents or AI tools creating 70% of new databases. These company metrics show software creation accelerating into an infrastructure problem.

Turso founder Glauber Costa will join as Head of Agentic Services alongside co-founder Pekka Enberg and the Turso team. Turso will keep operating, with a path from lightweight SQLite workloads into Supabase's Postgres ecosystem.

The financing follows Supabase's $500M June Series F, also led by GIC, at a $10B pre-money valuation. Supabase has not attached a new valuation to this $150M or disclosed how much is primary capital versus employee liquidity. The clean number is the announced financing, not an invented operating-cash total.

The strategic bet is sharper than the headline. AI agents may make database creation ordinary, but they do not make isolation, reliability, observability, cost control, or data governance optional. Every easy provision creates something a human must secure, understand, and graduate into production if it survives.

Supabase and Turso want to own that graduation path. That is a product promise with an infrastructure bill attached. Can they make a database as easy to create as a file without making its obligations invisible?

PaleBlueDot AI

The cleanest AI infrastructure product is usually held together by a messy promise: when the customer needs compute, the capacity has to exist in the right place, with the right hardware, and behave like one service.

PaleBlueDot AI is financing that promise at a new scale. The Palo Alto company has raised a $200M Series C led by ComputeCore at a $3.2B valuation. B Capital joined as an existing shareholder; other participants were not named.

Founders Jonathan Zhu and Sheldon Huang built a business that spans dedicated GPU clusters, a marketplace for partner capacity, and serverless inference. CEO Stephen Watts now has to make those 3 operating models feel boring to customers. Boring means the cluster arrives, the network holds, the scheduler works, the location fits, and a frontier lab or enterprise can run without learning which supplier owns which headache.

The demand signal is large. PaleBlueDot AI says it had signed more than $5B in customer contracts by the end of September, with U.S. and Japanese customers contributing more than half of monthly revenue. That contract figure is company-reported, not an audited revenue number. The company's website also reports 150+ clusters, 220,000+ GPUs connected, 60+ regions, and 25+ supply partners. “Connected” matters because the platform blends owned infrastructure with marketplace supply.

The company has been stacking capital almost as quickly as it stacks compute. A $150M Series B in January valued PaleBlueDot AI above $1B. In July, Brookfield Asset Management and Tor Investment Management provided a separate $255M 3-year private note. This Series C is equity, and the credit facility is not part of the $200M round.

There is technical evidence behind the pitch. PaleBlueDot AI says its Japan B300 cluster earned NVIDIA Exemplar Cloud status after exceeding 98% of NVIDIA reference performance across 6 training configurations and completing a week-long full-load test. That validation helps buyers compare a live system with a reference standard. It still leaves the commercial job where it belongs: turning financed capacity into reliable workloads.

The new capital will add compute capacity, broaden hardware and location choices, and support full-stack and go-to-market investment. That is less a victory lap than a bigger calendar. GPUs, data-center power, networking, storage, software, customers, and financing all have to meet on schedule.

The customer sees one platform. PaleBlueDot AI now carries the work of making every handoff behind it disappear.

AlgoX2

A benchmark can make infrastructure look inevitable. A production incident asks whether anyone should have trusted it.

AlgoX2 has raised a $10M seed extension to work on the part between those 2 sentences. Bessemer Venture Partners and Fin Capital led the round, with Alumni Ventures and Somersault Ventures joining alongside existing investors Octave, Motivate Venture Capital, and AAL VC.

The New York company was built by Alexei Lebedev, CEO, Vladimir Parizhsky, CTO, and George Levin, CBO. Lebedev and Parizhsky helped architect NYSE Pillar after careers spent building systems where delayed or disordered data becomes an operating problem, not a mildly annoying dashboard refresh.

Their new argument is that enterprise streaming has accumulated too many separate systems. Transport sits in one place, processing in another, state somewhere else, and analytics waits for its own copy. AlgoX2 puts streaming, processing, and storage on one data plane, with native support for Kafka, Redis, NATS, MQTT, HTTP, and WebSocket.

That architecture carries ambitious company-reported numbers: sub-10-microsecond publish acknowledgements, up to 5M messages per second per node, 10x fewer servers, and up to 90% lower total cost of ownership. Those claims still need the customer evidence that only production can supply.

AlgoX2 says the platform has been tested in production with unnamed design partners. The money is going toward forward-deployed engineers who can take integrations live, plus a product team led by former Confluent Cloud group product manager Addison Higham to launch AlgoX2 Cloud and an open-source version.

That use of capital is the interesting part. Deep infrastructure companies rarely lose because the demo was insufficiently clever. They lose in the gap between architecture and adoption, where somebody must map the customer's systems, survive the migration, explain failure modes, and be present when a workflow that looked clean on a diagram meets 12 years of enterprise history.

The founders already know how to build for unforgiving systems. This round buys AlgoX2 the chance to prove that experience travels from an exchange environment into finance, AI, and the broader enterprise without asking every customer to become a streaming specialist first.

If that handoff works, the company's best technical achievement may be making a sophisticated data plane feel uneventful on the day a customer finally depends on it.

Tangent Robotics

A robot can move the part across a factory and still lose the job at the thread, connector, gasket, or wire.

That final contact-rich step is where industrial automation gets humbled. A gripper can pick. A suction cup can place. Neither automatically knows how a component feels when it seats correctly, how much pressure is too much, or how to recover when the alignment is almost right and production needs it right.

Tangent Robotics has raised a $4.5M pre-seed round to work on that missing layer. Fly Ventures and Toyota Ventures co-led the financing, with Logos Fund and Sparked Ventures participating.

The New York company came out of Columbia Engineering's Robotic Manipulation and Mobility Lab. Co-founders Pedro Piacenza, CEO, Matei Ciocarlie, Chief Science Officer, and Ioannis Kymissis, a consultant for unconventional electronics, bring years of work across robot hands, optical touch sensing, and manipulation. Tianjian Chen is CTO.

The technical choice is more interesting than another attempt to manufacture a metal copy of the human hand. Tangent is combining custom articulation, light-based sensors inside the fingers, teleoperation and data collection, and machine learning into what it calls a fine-motor-skill layer. High-level models can decide what should happen. Tangent wants the hand to manage the delicate physical exchange that makes the decision useful.

Columbia documented tactile fingers that localized contact at sub-millimeter precision across curved surfaces, then demonstrated in-hand manipulation without external cameras. The lab also developed mechanisms for NASA's Astrobee robots, and TIME included its tactile hand among the Best Inventions of 2023.

The commercial proof is still ahead. Tangent has not disclosed customers, revenue, valuation, production deployments, or audited task-performance metrics. Its goal is to reduce practice for new skills to days and eventually hours. That remains a target.

The round gives the team room to move from research and public prototypes toward manufacturing. Assembly lines care about repeatability, cycle time, maintenance, recovery, and whether a machine can finish the connector without creating a quality problem three stations later.

That is why this pre-seed matters beyond the amount. Physical AI will not earn its place on the factory floor because a robot can reason about the task. It will earn it when sensing, mechanics, and learning survive the moment the part touches the world.

Biotia

A diagnostic can find what standard testing missed and still struggle to become standard care.

Biotia raised an oversubscribed $9M financing to work on that second problem. The New York company uses clinical metagenomics and machine learning to identify pathogens behind challenging infections. The capital will support laboratory capacity, reimbursement, new women's-health diagnostics, and an expansion into orthopedics.

CEO Niamh O'Hara, PhD, and Global Director Christopher Mason, PhD, co-founded Biotia from research at Cornell Tech. CTO David Danko, PhD, leads the software and bioinformatics work behind its diagnostic pipelines.

BIOTIA-ID Urine is designed for recurrent and complicated UTIs, where standard culture can leave symptoms without a useful explanation. After comparator adjudication, a 2026 peer-reviewed study of 1,470 urine specimens and 14,500+ analytes reported 97.2% sensitivity and 99.6% specificity.

Those numbers deserve attention, but a diagnostic business is never built on sensitivity and specificity alone. The laboratory must handle more volume. Clinicians need to know when to order the test and how to use the result. Patients need access. Claims need a path through reimbursement. A technically better answer can still lose to a familiar workflow.

Biotia's funding plan reads like a company moving into that commercial reality. The capital will scale its high-complexity, CLIA-certified New York City laboratory, expand its test menu, pursue reimbursement, and support additional diagnostics in women's health. The company is also developing a research-use-only joint-fluid test with Hospital for Special Surgery as it moves toward orthopedic infections.

Tim Barry and Clive Fields joined Convergent Ventures, DigitalDx Ventures, Continuum Health Ventures, Leawood Venture Capital, and other participants. Biotia did not disclose a lead, round label, valuation, or detailed terms.

Biotia previously raised an $8M Series A led by OCA Ventures in 2022. It has since secured New York State CLEP approval for its urine test, expanded availability nationwide, added drug-resistance reporting, received a test-specific AMA billing code, and announced collaborations with Mayo Clinic and Hospital for Special Surgery.

That sequence explains why this $9M matters. Biotia is no longer financing only a scientific possibility. It is financing the handoffs that determine whether clinical metagenomics becomes useful at the exact moment a physician and patient need an answer.

NorthStar Medical Technologies

An actinium-225 therapy starts losing usable material before it reaches a patient. With a roughly 10-day half-life, every production, testing, release, manufacturing, and shipping delay spends inventory nobody can get back.

NorthStar Medical Technologies has secured a multi-year growth-capital facility of up to $185M from Hercules Capital to make that clock less expensive. The company said $100M is being provided upfront to expand isotope production and clinical and commercial radiopharmaceutical manufacturing at its Beloit, Wisconsin campus.

The accounting matters. NorthStar has access to a facility worth up to $185M; it did not say the full amount arrived at closing. Pricing, maturity, future draw conditions, valuation, and ownership implications were not disclosed.

CEO Frank Scholz is building around a simple operating idea: put the isotope plant and drug-manufacturing work on the same campus. NorthStar produces no-carrier-added Ac-225 with electron-accelerator technology. Its CDMO supports clinical programs, cGMP manufacturing, analytical work, packaging, and radioactive-material logistics.

That makes vertical integration a literal race against decay. A biotech needs more than an isotope supplier. It needs purified material, quality release, manufacturing capacity, regulatory documentation, and delivery to line up before too much useful material disappears.

NorthStar entered 2026 with the science moving into industrial territory. In January, it announced commercial-scale production of n.c.a. Ac-225. In April, FDA accepted the company's Type II Drug Master File, giving partners a regulatory file they can reference in their own submissions. That is development infrastructure, not approval of a therapy.

The company reports more than 30 pharmaceutical and biotechnology supply and manufacturing agreements and plans a higher-capacity Ac-225 line later this year. Exact output, revenue, customer economics, and customer names remain undisclosed.

Co-founder and CSO James Harvey brings continuity from NorthStar's earlier medical-isotope work. CTO Jeff Schaaf carries the expansion across operations, quality, CDMO development, IT, and construction. The financing puts more weight on the handoff between technical depth and repeatable manufacturing.

For NorthStar's customers, the product is not simply a vial of Ac-225. It is usable time. The company now has $100M upfront, and a larger facility behind it, to make more of that time survive the trip through Beloit.

Arceus

The billable hour has a strange way of turning delay into inventory.

A contract sits in a queue, sales waits, and every extra review cycle can become more revenue for the firm doing the reviewing. Most legal AI companies sell software into that system. Arceus is trying to own the outcome instead.

Arceus launched publicly with $17M in funding led by Greycroft, with Craft Ventures, South Park Commons, and others participating. Business Insider says the financing brought total funding to $17M, so the clean accounting is $17M in reported funding rather than a fully reconciled new tranche.

Founder and CEO Mac Liu is building an AI-native law firm, not another dashboard for an overloaded legal team. Licensed U.S.-barred attorneys handle every matter, while CounselOS carries context from client systems, past agreements, preferences, and negotiation history.

That moves the automation inside the service. The client sends the contract and gets legal work back. Arceus says reviews average 3-5 hours, guarantees an 8-hour turnaround or the review is free, and prices matters upfront instead of billing by the hour.

The company reports reviewing more than 12,000 contracts, reducing legal spend by more than 70%, and quadrupling its customer base in September. Those are unaudited company claims, but they make the logic visible: speed becomes the product only when an attorney still owns the judgment.

Chief Legal Officer Eugenia Bergantz leads the legal function. The capital is meant to add attorneys, engineers, and AI researchers while expanding from contract review into negotiation, employment, privacy, and ongoing counsel.

This is the more consequential legal AI wager. Software vendors ask firms to use automation inside economics still tied to hours. Arceus changes the economics first, then uses automation to make flat-fee work faster and more predictable.

That also moves the risk. The firm, not the customer, has to make the AI useful without letting speed outrun legal judgment. Contract review is a disciplined wedge. Employment, privacy, and ongoing counsel will show whether CounselOS can carry institutional memory into work where context is heavier and the answer is less repeatable.

Greycroft partner Mark Terbeek framed the opportunity as rebuilding the law firm itself. The $17M buys room to prove whether a legal service can return work in hours and still put a licensed professional behind the result when the contract leaves the queue and reaches the business.