The price of a plastic resin is a number, but the cost is a story. It’s a story about which cracker it came from, what that plant paid for its gas last month, how many tankers are queued at the port, and how many tons of CO2 were emitted to make it. For decades, procurement teams at large manufacturers have bought that story from their suppliers. Atmospheric AI is betting they would rather compute it themselves.
The London based startup, founded in 2025, builds computational models of industrial supply chains, facility by facility. Its product, a B2B analytics platform called Cortex, is designed to give procurement, sustainability, and R&D teams an independent view of what raw materials should cost, where capacity truly exists, and what carbon is baked into their sourcing decisions [Axios Pro, November 2026]. The company raised a $10.1 million seed round in November 2026, led by climate tech focused investor DBL Partners, with participation from Raise, Broadscale, and others [Axios Pro, November 2026].
The wedge is independent ground truth
Atmospheric AI’s initial focus is the petrochemical and plastics supply chain, a sprawling global network of crackers, refineries, and shipping lanes. The company says it has modeled more than 50,000 industrial facilities worldwide, incorporating data on their technologies, efficiencies, and energy mixes to track material flows [Axios Pro, November 2026]. The core proposition is moving from supplier provided estimates to a computed, facility level view of cost, risk, and emissions.
This isn’t just another ESG reporting dashboard. The company combines chemical engineering, plant economics, industrial process data, and AI to model the actual production of raw materials [The SaaS News, November 2026]. For a manufacturer sourcing a specific polymer, the platform can theoretically show the marginal cost of production at different plants, the carbon intensity of each route, and how a drought or geopolitical event might ripple through the network. The goal is to turn sourcing from a negotiation into an optimization problem.
A team built for enterprise credibility
The founders have assembled a team that speaks the language of both large scale software and corporate carbon transition. Co founder and CEO Roee Goldberg previously co founded OpenWeb, a community engagement company he helped scale [Axios Pro, November 2026]. The more telling hire for the company’s ambitions is Torsten Lichtenau, the founding chief commercial officer, who spent 22 years at Bain & Company, most recently as a senior partner and global head of its carbon transition practice [nstartup.me, November 2026]. Lichtenau didn’t leave Bain’s partnership to sell a simple dashboard; he left to sell a new kind of operational intelligence.
The investor table also signals a focus on tangible industrial and climate impact. DBL Partners, the lead, is known for backing companies like Tesla and SpaceX in their early days, with a thesis around both financial returns and positive environmental impact. DBL’s Vinnie Caruso is joining the company’s board [tradepoint.io, November 2026].
| Role | Name | Key Background |
|---|---|---|
| Co founder & CEO | Roee Goldberg | Former co founder and COO of OpenWeb |
| Founding CCO | Torsten Lichtenau | Former Bain & Company senior partner, global head of carbon transition |
| Board Member | Vinnie Caruso | Partner, DBL Partners |
Early traction with manufacturers
The company is working with an initial group of approximately 100 manufacturers, according to a November 2026 report [The SaaS News, November 2026]. Its most cited early customer is consumer health giant Haleon, which reportedly worked with Atmospheric AI for almost two years on resin sourcing by the time of the funding announcement [Startup Mag, November 2026]. A two year engagement before a formal seed round suggests a product built in close collaboration with a demanding first customer, not in a vacuum.
The company’s public job postings hint at the technical scope. It is hiring senior full stack engineers for roles described as "AI Powered Procurement," indicating the platform aims to be more than a static reporting tool [freehire.me]. Other product descriptions mention an AI assistant for conversational access to engineering data and a "Context Graph" to map relationships automatically [GeniusFirms, October 2026].
The expansion and the obstacles
The roadmap is clear: prove the model in plastics, then expand. The company plans to move into adjacent raw material verticals like chemicals, paper, and metals [The SaaS News, November 2026]. The $10.1 million in seed capital is earmarked for expanding these models and supporting commercial hiring [The SaaS News, November 2026].
The bet is compelling, but the path is lined with credible challenges. The company must convince cost focused procurement departments that this intelligence is worth a premium SaaS fee, on top of the millions they already spend on materials. It must continually ingest and validate vast amounts of disparate industrial data to keep its digital twins accurate. And it must expand without diluting the deep, vertical specificity that makes its plastics model valuable.
- Data integrity. The model is only as good as its inputs. Sourcing reliable, timely data on 50,000 facilities’ operational efficiencies and energy sources is a perpetual challenge.
- Procurement inertia. Shifting billion dollar sourcing decisions away from longstanding supplier relationships requires proving undeniable ROI, not just interesting analytics.
- Vertical expansion. Success in petrochemicals does not automatically translate to paper or metals, each with its own unique processes, players, and data challenges.
The company’s answer likely lies in the depth of its models and the credibility of its team. A Bain veteran selling a platform that embeds carbon cost directly into sourcing conversations is a different proposition than a generic analytics startup.
The unit economics of a ton of CO2
Let’s put a pencil to it. If a large manufacturer sources 100,000 tons of a particular resin annually, and Atmospheric AI’s model identifies a sourcing mix that is 5% cheaper and 10% lower carbon, the savings are material in both senses. At a resin price of $1,500 per ton, the 5% cost saving is $7.5 million per year. The carbon saving, perhaps 15,000 tons of CO2, is harder to price but increasingly material on corporate balance sheets. The platform’s value isn’t in pointing out that carbon exists, it’s in showing the cheapest way to avoid it. For Atmospheric AI to succeed, it doesn’t need to beat a flashy AI competitor. It needs to beat the incumbent: the spreadsheet, the supplier’s word, and the comfortable opacity of the traditional supply chain.
Sources
- [Axios Pro, November 2026] Exclusive: Atmospheric AI raises $10M to track materials supply chain | https://www.axios.com/pro/climate-deals/2026/11/04/atmospheric-ai-dbl-seed
- [The SaaS News, November 2026] Atmospheric AI Raises $10M Seed | https://www.thesaasnews.com/news/atmospheric-ai-raises-10m-seed/
- [Startup Mag, November 2026] Atmospheric AI raises £7m from DBL Partners and RAISE | https://www.startupmag.co.uk/funding/atmospheric-ai-2026-seed-funding/
- [nstartup.me, November 2026] Atmospheric AI Raises £7M to Map Industrial Supply Chains | https://nstartup.me/news/4e73b8c5-4a3d-4ff9-affd-54a70d8fa04c
- [tradepoint.io, November 2026] Atmospheric AI Secures $10M Seed for Procurement Modeling Push | https://tradepoint.io/atmospheric-ai-secures-10m-seed-for-procurement-modeling-push-unite-ai/
- [freehire.me, retrieved 2026] Atmospheric AI company page and job postings | https://freehire.me/companies/atmospheric-ai
- [GeniusFirms, October 2026] Cortex platform description | https://geniusfirms.com/product/cortex-2
- [Martin Svensson - LinkedIn, retrieved 2026] Post on Atmospheric AI's facility modeling | https://www.linkedin.com/in/martinsvensson
- [Danny Alexandrovitz - LinkedIn, retrieved 2026] Post on Torsten Lichtenau's role | https://www.linkedin.com/in/danny-alexandrovitz-9654b9108/
- [Chris Morton - LinkedIn, retrieved 2026] Post on Torsten Lichtenau's move from Bain | https://www.linkedin.com/in/chris-morton
- [Markus Heinen - LinkedIn, retrieved 2026] Post on Vinnie Caruso joining the board | https://www.linkedin.com/in/markus-heinen
- [Sam Isaacs - LinkedIn, retrieved 2026] Post on Roee Goldberg's background | https://www.linkedin.com/in/sam-isaacs