# Michael Freedman

Michael Freedman works on data infrastructure for industrial and AI systems at Tiger Data, the company behind TimescaleDB. He writes and speaks about connecting industrial control systems and machine data with PostgreSQL for real-time operations and historical analysis.

- Profile: https://b2binfluencer.marketing/creators/michael-freedman/
- LinkedIn: https://www.linkedin.com/in/mfreed
- LinkedIn followers: 8,028
- Topics: artificial-intelligence
- Based on 13 dated posts, latest 2026-09-24

## Summary

They are most useful for technical content about industrial control, PostgreSQL, and time-series data for AI systems.

- Best for: A data infrastructure company seeking a technical partnership or content collaboration on industrial AI systems.
- Not for: A consumer AI writing assistant wanting paid LinkedIn sponsorship with discount code and general AI productivity message.
- Check first: Confirm recent results and audience demographics for comparable industrial, manufacturing, or time-series work.

### Takeaways

- Power is becoming the binding constraint on AI data centers, not just an optimization variable.
- Industrial control and Postgres belong together because machine data needs history and relational context.
- Titan America uses TimescaleDB for SCADA readings, dashboards, and 19 TB of time-series data.

## Main ideas

- **Power is the binding constraint on AI data centers, not an optimization variable**: Freedman’s argument is that, for years, compute treated power as something to optimize; now power itself is the constraint. AI data center challenges are moving out of the facilities department and into systems architecture: power, cooling, water, grid interconnects, backup generation, and the ability to run dense compute reliably set the digital performance envelope. Operational details such as power, cooling, capacity, utilization, and equipment state are no longer just dashboard metrics; they’re part of the architecture. They also describe this as a real-time control problem as power and heat density rise and more facilities move toward liquid cooling.
- **Industrial control and Postgres belong in one system because machine data needs history and relational context together**: Freedman’s work centers on pairing industrial control software with Postgres that can scale to the volume and history of the physical world. Their reasoning: real-time data alone isn’t enough; systems also need historical data combined with the relational context that gives it meaning. They frame this as an architecture question—how the stack got here and where it goes next—aimed at controls engineers and system integrators as much as software buyers.
- **Customer proof should show specific operational scale, not a generic testimonial**: Freedman uses named industrial customers and hard operational figures as evidence. The point is to show machine data running at real volume, with the database layer supporting both real-time dashboards and business reporting while managing storage tiers. The concrete opening—much of the modern world is built on concrete, and some of that concrete runs on Postgres—signals that heavy industry is the intended proof, not a software-only reference.
- **AI is driving a re-industrialization wave that needs machine, sensor, and operations data infrastructure**: Freedman’s commercial argument is that AI isn’t only changing software; it’s driving a re-industrialization wave that needs modern data infrastructure for machines, sensors, operations, and real-time intelligence. Demand is coming from physical-industry customers, not only software, making industrial data infrastructure a growth thesis rather than a niche.

## Work to read

- [Ignition Community Conference: industrial control meets Postgres](https://www.linkedin.com/posts/mfreed_im-talking-at-the-ignition-community-conference-activity-7504250356319727617-1aQ3)
- [Why power, not chips, is becoming the constraint on AI data centers](https://www.linkedin.com/posts/mfreed_for-years-compute-treated-power-as-something-activity-7465439458100736001-CUh5)
- [Titan America: SCADA readings and 19 TB of time-series data on Postgres](https://www.linkedin.com/posts/mfreed_how-titan-america-unified-manufacturing-sensor-activity-7506068458627862528-C-Od)
- [Announcing the Ignition Enterprise Historian on stage at the Ignition Community Conference](https://www.linkedin.com/posts/mfreed_iccunleashed-activity-7508905589167734786-wBkj)
- [What a quarter of roughly double expected revenue says about industrial data demand](https://www.linkedin.com/posts/mfreed_what-a-quarter-had-2x-expected-revenue-activity-7462599332920283137-QJTA)

## Brand relationships

- **Tiger Data** (employment)
- **Inductive Automation** (disclosed_partnership)
- **Titan America** (client)

## Partnerships

Michael Freedman has no published sponsorship policy, rate card, or open partnership application. The collaborations mentioned are product and platform collaborations, not paid placements: Tiger Data's work with Inductive Automation was announced jointly, and Freedman appeared onstage with Inductive Automation's CEO to launch a first-party product built on TimescaleDB Enterprise. A brand approaching them should frame the ask as a technical partnership or content collaboration tied to industrial data infrastructure, since every collaboration mentioned centers on an integration or shared technical story rather than a sponsorship slot. No official inquiry route is named, so ask for the appropriate contact directly.

## Alternatives

- Simon Allen: Choose Simon Allen when the collaboration is about market-level positioning and convening rather than product integration. Allen's posts show them working the UK digital infrastructure agenda directly: representing the Quantum Infrastructure Council at Platform UK, noting that every discussion stayed grounded in UK realities such as planning reform, grid constraints, regional growth, investment appetite, supply chain delivery and skills, and moderating a Capacity Europe panel titled 'Scaling Under Scarcity: Delivering Projects Amid Shortages' covering power equipment, chips, cooling, supply chains and skilled resources. A data infrastructure company seeking a co-hosted panel, a market-strategy article or an introduction to UK data centre and energy audiences would get a more direct route here than through a technical partnership post. Allen also frames quantum readiness as an infrastructure question tied to power, cooling, site design and long-term strategy, which suits a buyer focused on planning and delivery constraints rather than the data layer itself.
- Trent Warren, MBA, TEFL: Choose Trent Warren when a campaign needs supplier-side market intelligence rather than deployment experience. The posts use proprietary data to answer quantitative questions about AI infrastructure supply and demand: why average memory per hyperscaler server exceeds roughly 600 GB in regions like Colombia while the U.S. stays below 400 GB, and why average AWS spot prices rose about 14.3% over three months as generative AI workloads absorbed spare compute and reserved blocks shrank the spot pool. A company whose campaign angle is 'what hyperscaler capacity and pricing data tells buyers about AI demand' would get a quantitative, data-led voice here. Both posts close by inviting people to talk about seeing more global data across cloud service providers, AI and semiconductors, pointing to a data-subscription or market-intelligence relationship.
- Joe Yan: Choose Joe Yan when the collaboration is an investment or market thesis about physical AI and the industrial supply chain, not a data-platform integration. The posts map how AI demand propagates into physical suppliers: Generac, founded in 1959, now building large-scale data centre power systems with an approximately $1.6 billion data-centre backlog, and Modine HVAC, founded in 1916, supplying mission-critical cooling through its Airedale business under a long-term agreement worth more than $4 billion between 2027 and 2029, with $165 million paid upfront to fund capacity. Yan also lays out a China-U.S.-Japan thesis on embodied AI, arguing the next phase of AI investing is less about separating software from hardware and more about understanding how software demand moves through the physical supply chain. A buyer whose message is 'the beneficiaries of AI may not look like AI companies' would get a clear, portfolio-level framing here, plus a comparative read on where physical AI is funded and deployed.

---
AI-assisted summary of public posts. Audience demographics are not available. Engagement appears only where we checked recent posts. Method: https://b2binfluencer.marketing/methodology/