TUBR's Physics-Based AI Takes On Demand Forecasting With Less Data

The Sheffield startup, backed by £154,000, is betting its physics-based machine learning can forecast demand for cafes and grocers where conventional AI needs more data.

About TUBR

Published

Most demand-forecasting software asks for more data than a small cafe or independent grocery store could ever provide. TUBR, a Sheffield-based startup, is building its product on the opposite premise. Its platform, TUBR Pulse, is designed to generate operational forecasts for food retailers using what the company calls a physics-based machine learning engine, tailored for businesses with limited historical data [BusinessLive, November 2025].

Founded in 2020 by Dash Tabor, the company recently secured £154,000 in angel funding to expand its technical team and launch commercial trials, with a notable pilot program set for Phoenix, Arizona [BusinessLive, November 2025] [Metro Phoenix Alliance, March 2026]. For a sector where waste and staffing are daily margin battles, the promise is clear. The harder question is whether a novel technical approach can consistently beat established rules of thumb, and at what price.

The wedge of small data

TUBR's core bet is that its methodology can deliver accurate predictions where others see noise. Conventional machine learning models for demand forecasting often require extensive, clean datasets to train effectively,a bar many small and medium-sized businesses cannot clear. TUBR positions its physics-based engine as a differentiator, applying principles that can work with sparser operational inputs to forecast sales, staffing needs, and potential waste [Sheffield Digital, May 2023].

The product surfaces are pragmatic, targeting the specific pain points of restaurant and grocery managers. TUBR Pulse aims to help users predict variable costs, identify which operational levers to pull in real-time, and plan supply chains more efficiently [BusinessLive, November 2025]. It's a classic vertical software play, but one hinging on a technical moat rather than just a better dashboard.

Funding and founder resilience

The recent £154,000 angel round was led by Lifted Ventures, a UK network focused on backing female founders outside of London [BusinessLive, November 2025]. The capital is earmarked for team growth and pilot programs, a logical use of proceeds for a company at this stage. The funding story, however, is underscored by a narrative of founder resilience that has become part of the company's early identity. Dash Tabor has publicly discussed rebuilding the team after losing a co-founder during a fundraise and facing significant financial setbacks during development [The Everyday Founder, November 2025].

While the founding team listed in some records includes Evanalison Tabor and Nikhil Kanta, the public narrative centers on Dash Tabor as the driving CEO [Tracxn]. Nikhil Kanta, described as an entrepreneur and physicist, served as CTO until May 2023, lending credence to the technical origins of the "physics-based" claim [NIKHIL N KANTA Resume, Retrieved 2026].

Traction and the Phoenix pilot

Measurable traction for a deep-tech startup at seed stage is often about signals, not scale. TUBR's most concrete signal is its planned commercial trial in Phoenix, marking a strategic transatlantic entry into a new market [Metro Phoenix Alliance, March 2026]. The company has also stated it secured customers earlier in its development, though it has not named them publicly [Sheffield Digital, May 2023].

The current competitive landscape for retail forecasting is stratified, which clarifies TUBR's potential lane.

Competitor Primary Focus Typical Customer Size
TUBR Pulse Predictive analytics for cafes, restaurants, grocers Small to Medium Businesses (SMBs)
o9 Solutions Enterprise supply-chain planning & analytics Large Enterprise
SymphonyAI AI-powered vertical solutions for retail & CPG Mid-Market to Enterprise
Perimattic Demand OS Demand forecasting & planning platform Mid-Market to Enterprise

As the table shows, TUBR is aiming at a segment,SMB hospitality,that the larger platform players often underserve due to cost and complexity. Its competition is as much spreadsheets and intuition as it is other software vendors.

The risks in the recipe

The ambition is clear, but the path to sustainable enterprise value is lined with specific challenges any procurement officer would note. The company's technical differentiation, while compelling, remains largely unproven at commercial scale. The success of the Phoenix pilot will be a critical first validation of both the model's accuracy and its operational fit [Metro Phoenix Alliance, March 2026].

Furthermore, the business model for selling deep-tech AI to small businesses is notoriously difficult. Sales cycles can be long, acquisition costs high, and price sensitivity extreme. TUBR will need to demonstrate not just predictive accuracy, but a tangible ROI that justifies its yet-to-be-disclosed subscription fee. The company's answer appears to be a focus on very specific, high-value outcomes like waste reduction and labor optimization, where savings can be directly measured [BusinessLive, November 2025].

The ideal customer and the road ahead

TUBR's ideal customer profile is not a generic retailer. It's the owner of a small chain of cafes who knows that throwing away unsold food every afternoon is a silent profit killer, or the manager of an independent grocery store who needs to schedule staff without the luxury of a dedicated analyst. This is a buyer who feels the pain of operational guesswork daily but lacks the data infrastructure of a large chain [BusinessLive, November 2025].

The realistic competitive set extends beyond the named software platforms. For this ICP, the true alternative is often a veteran manager's intuition, a basic spreadsheet, or a point-of-sale system's rudimentary reports. To win, TUBR must be significantly better than that status quo, and simple enough to adopt without a data science team.

The next twelve months will be defining. Success will be measured by the results of the Phoenix pilot, the conversion of those trials into paying contracts, and the subsequent expansion of the team hinted at with the recent funding. For Dash Tabor and TUBR, the bet is that in the messy, data-sparse world of small business hospitality, a little physics can go a long way.

Sources

  1. [BusinessLive, November 2025] Yorkshire innovator TUBR to ramp up work after sealing six-figure investment | https://www.business-live.co.uk/enterprise/yorkshire-innovator-tubr-ramp-up-32947020
  2. [Sheffield Digital, May 2023] How did you get here: Interview with Founder & CEO of AI machine learning startup TUBR - Dash Tabor | https://sheffield.digital/podcast/how-did-you-get-here-interview-with-founder-ceo-of-ai-machine-learning-startup-tubr-dash-tabor/
  3. [The Everyday Founder, November 2025] How She Turned a Crowded Tube Ride into a Deep Tech Startup (and Nearly Lost It All) | Dash Tabor | https://podcasts.apple.com/gb/podcast/how-she-turned-a-crowded-tube-ride-into-a-deep/id1732914477?i=1000737548942
  4. [Metro Phoenix Alliance, March 2026] UK-based AI Startup Launching Analytics Platform in Phoenix | https://metrophoenix.com/2026/03/uk-based-ai-startup-launching-analytics-platform-in-phoenix/
  5. [Tracxn] TUBR - 2026 Company Profile, Team, Funding, Competitors & Financials - Tracxn | https://tracxn.com/d/companies/tubr/__QxJxwGzhUJO7agHlZ_OdPsnIGEGvSWErljRdtkB0-v4
  6. [NIKHIL N KANTA Resume, Retrieved 2026] NIKHIL N KANTA Resume | (Source provided in research snippets)

Read on Startuply.vc