Helping companies operate at their full potential.
AI has already shown how dramatically it can increase the productivity of an individual. We believe it can do the same for entire companies.
Rithmo exists to remove the organizational friction that keeps businesses from capturing that potential, so that people and AI work from the same current, trusted understanding of the business.
Two products taught us the same lesson.
We did not start with a thesis about context. We started with a question about productivity, and got here by building things that ran into the same wall twice.
December 2024
Firmware written by someone who had never written firmware.
Tim built his first working firmware with ChatGPT, having never been a software engineer. The tooling was clunky, context ran out constantly, and hallucinations were routine. Eventually it compiled. A capability constraint that had shaped his entire career had stopped being one.
December 2025
Rithmo was founded around the question that followed.
If AI could make one person that much more capable, could it do the same for an entire company? Meetings were the most obvious place organizational capacity went missing, so we started with an AI meeting host.
The first wall
We built the technology. The market showed us its limit.
Making meetings meaningfully more efficient meant understanding and influencing how people behave. People were impressed by what we had built, and they did not want AI analyzing or characterizing their behavior.
What we found there
Keeping decisions accurate is extraordinarily hard.
Extracting a sentence that sounds like a decision is easy. Understanding what was actually decided, carrying that across separate conversations, tying it to the evidence behind it, tracking how it changed, and recognizing when one decision quietly replaced another is a different problem entirely.
The second wall
We moved one layer down and met the same resistance.
We explored decision intelligence as the product itself. A system that surfaces dropped, stalled, or poorly handled decisions also makes individuals feel scrutinized. The capability was valuable. Packaging it around human accountability created defensiveness.
At the same time
Agents were getting capable, and a human was holding them up.
Individual agents worked largely because a person managed their context continuously: choosing the right document, correcting stale information, updating instructions, resolving confusion, pointing the agent at the current version of reality. That does not scale across a company of documents, systems, exceptions, owners, and decisions all changing at different speeds.
The decision infrastructure we had been learning to build was never primarily a system for monitoring people. It is what AI agents need in order to operate safely inside a real company.
Which put us back on the question we started with: letting an entire organization capture the productivity that AI already gives one person.
Enterprise systems hold what was written down, not what is true today.
Documents, CRM fields, tickets, policies, contracts, and schemas are authored records. They matter, and they are not the whole picture. A business runs continuously: people decide, change direction, create exceptions, resolve ambiguity, and supersede what used to be true.
An agent cannot safely operate from the recorded half alone. Rithmo reconciles the recorded side of a business against its current resolved state.
Across the company
Rithmo connects context across systems and conversations that do not talk to each other, so no agent has to work from whichever slice it happens to be pointed at.
Current, not just recorded
Rithmo maintains what the business has actually decided, including what changed, who owns it, what evidence it rests on, and what it superseded.
Ready for AI action
Rithmo gives agents a current, trusted state to check against before they act, and lets them stop when the business has not actually resolved something.
A track record that predates Rithmo.
Built & sold
NanoThings
Founded 2016 · Acquired 2024
Tim founded the company and invented the NanoTag, the first smart label of its kind, a LoRa device with pharmaceutical-grade temperature monitoring and RSSI location tracking. Its insight was to keep the tag itself location-unaware and resolve position in the cloud from raw signal, so it needed no GPS or Wi-Fi and ran at a fraction of the cost and power. It was the first commercial device to turn raw RSSI into usable location, on a mechanically flexible, waterproof circuit thin enough to live on paper, cardboard, or plastic and cheap enough to throw away.
He built the manufacturing for it too: a hybrid printed-electronics line that fused two industries that had never met, commercial printing and semiconductor pick-and-place assembly, to produce a complex, fully functional device from a circuit printed on paper or polyimide film, something neither had done before. The hard part was never the software but the embedded system, firmware, and RF protocol; he took the LoRaWAN stack, broke it, and rebuilt it for the device, the same move he would later repeat on the wireless protocol behind the DoD system.
Systems architecture
Defense and life sciences
Tim architected a precision indoor asset-localization system deployed in U.S. Department of Defense environments, built on a custom multi-tier wireless protocol spanning edge devices, gateways, and cloud infrastructure. At its core was a neural network that resolved position from raw RSSI signal, learning location out of noisy radio where fixed math could not. It was AIoT before the term existed.
For life sciences he designed a mechanically flexible pharmaceutical smart label with real-time location awareness and two bidirectional radios, Bluetooth and LoRaWAN, certified to fly aboard commercial aircraft. Its onboard protocol adapted in real time, detecting flight, switching to store-and-forward, and continuously tuning power draw, memory, transmit strength, and retries to last a full journey on a label-sized battery. It is now used across major life-sciences supply chains.
Venture & governance
Investor, board member, and fund GP
Raghu spent more than a decade as an investor and board member across AI, enterprise software, IoT, and digital infrastructure. Companies he backed were acquired by Microsoft, Web.com, and YipitData, and another went public. He then became General Partner of a pre-seed fund, building a 22-company portfolio and the operating and governance systems that held it accountable.
That vantage is why he is building Rithmo. From a board seat, across dozens of companies at once, he watched how decisions actually get made, who owns them, and how quietly they get lost, which is the exact failure Rithmo exists to fix. At Rithmo, governance and operating discipline are not adjacent to the product; they are the product. Raghu spent a career installing that discipline in other people's companies, and now builds it into the layer that governs AI agents inside them.
The founders.
Tim Williams
Co-founder & CEO
Tim is a repeat founder and systems architect who has spent his career building complex technology products from the ground up.
He founded NanoThings in 2016, where he helped pioneer smart-label technology for cold-chain tracking and developed the NanoTag, a sticker-like wireless tracker and temperature monitor. NanoThings went through Plug and Play Silicon Valley in 2020 and was acquired in 2024.
Since then he has designed tracking and localization systems for U.S. Department of Defense and life-sciences applications, spanning smart labels, edge devices, gateways, wireless protocols, and cloud infrastructure.
Rithmo is the second company he has taken into Plug and Play, selected for its Enterprise AI accelerator in 2026.
Raghu Bharat
Co-founder & CTO
Raghu is a computer engineer by training and a technical founder who has spent his career building technology and working with companies from their earliest stages through major exits.
He began his career as an engineer at Nokia Siemens Networks, working on real-time signaling and handover behavior in carrier networks. He went on to spend more than a decade investing in and serving on the boards of technology companies across AI, enterprise software, IoT, and digital infrastructure. Companies he backed were later acquired by Microsoft, Web.com, and YipitData, and another completed a public offering.
He later became General Partner of a pre-seed venture fund, where he helped build a 22-company portfolio and the operating and governance systems behind it. He holds a BEng in Computer Science and Engineering and an MBA, both from McGill University.
What AI did for one person, at the scale of a company.
Companies should be able to put AI to work across their operations without those systems getting trapped by stale context, fragmented systems, bureaucracy, and the same information gaps that have always slowed organizations down.
Rithmo exists to help companies operate as close to their full potential as possible.