AKA Foods Launches AKA Studio Alchemy and Opens the Beta for AKA Atlas, a Private Knowledge Engine for Food

AKA Foods, the European food technology company behind AKA Studio, today announced two products for food R&D teams. AKA Studio Alchemy, the third generation of its secure AI platform, is available now to all customers. AKA Atlas, a private knowledge engine that builds and maintains a verified world model for any food domain, is opening its beta programme for companies that want early access.

Food companies hold decades of formulation history, sensory results, supplier data and regulatory knowledge. Most of it sits in spreadsheets, email threads and the heads of senior staff. When an R&D team starts a new project, they often start from scratch, not because the answers don't exist, but because nobody can find them fast enough. Alchemy and AKA Atlas tackle that problem from two directions: one does the formulation work, the other builds the knowledge to do it well.

AKA Studio Alchemy: from informed to agentic

Previous generations of AKA Studio helped food technologists find answers and organise data. Alchemy goes further: it acts. The platform now creates projects, generates batches of candidate recipes, reads product labels from photographs and updates development briefs. It always requests approval before taking action, and the expert always decides.

Alchemy is built around the development loop food teams already run: Ideate, Formulate, Sensory, Audit. Teams explore new product directions in Ideate. In Formulate, Alchemy generates and refines candidate recipes against layered constraints: cost caps, nutrient targets, ingredient restrictions. Sensory feedback from each tasting round feeds directly into the next iteration, with nutrition, cost and composition updating live with every edit. And at any point, a food technologist can ask Alchemy to run a regulatory Audit on the current formulation: it checks the relevant requirements, flags deviations, and shows which constraints held, with full sourcing transparency on every suggestion.

“We built Alchemy around the loop that food technologists already run: Ideate, Formulate, Sensory, Audit. Alchemy takes on the tasks inside that loop that don't need a scientist's palate, the searching, the constraint checking, the compliance tracking, and the scientist stays in charge of the science.” David Sack, CEO and co-founder, AKA Foods.

Alchemy arrives with 63 improvements across the platform. Recipe generation produces multiple candidates simultaneously, with controls that let teams choose between constrained options and more exploratory suggestions. The platform reads briefs in Word and Excel formats and generates downloadable reports including project summaries and QA/HACCP documentation. Teams can start working immediately and integrate their own data when ready.

AKA Atlas: the world model for food (beta)

AKA Atlas builds and maintains a private, verified body of knowledge for an organisation's domain: a model of its food science, computed, not retrieved. It connects everything a company knows, its own formulation history, sensory results, process learnings and supplier data, with everything published externally: scientific literature, patents, regulatory updates and ingredient specifications.

Ask AKA Atlas a question and it answers with evidence attached, sources cited and a confidence score. Where it finds gaps, it dispatches automated research scouts that locate, qualify and bring back verified information. When external evidence contradicts internal data, AKA Atlas surfaces both sides so teams make informed decisions rather than working from assumptions.

AKA Atlas is forkable. A single verified world model can be branched by geography or category: an ingredient company might maintain one for beverages, another for bakery and a third tuned to Brazilian regulations, all inheriting from a shared foundation and kept current by their own scouts. AKA Atlas connects above the systems a company already runs, through read-only connections, with nothing migrated and nothing replaced. It supports Bring Your Own Model, so companies can connect their own prediction models, and Sovereign AI deployment, so the reasoning engine can run on open-weight models inside the customer's own infrastructure.

“Every food company is using AI now. Almost none of them owns any intelligence. The chatbots their teams use are shared with competitors, ignorant of their business and forget everything the moment the window closes. AKA Atlas builds the private, verified knowledge that makes AI answers actually worth trusting.” David Sack, CEO and co-founder, AKA Foods.

“Most AI systems retrieve documents and summarise them. AKA Atlas is different: it holds a model of your domain, typed, connected and scored for confidence, so answers are computed from what your organisation actually knows, with the evidence attached. Nothing trains a shared model, and your data never leaves your boundary.” Professor Alex Bronstein, Chief Scientist, AKA Foods.

Private by design

Both products operate within AKA Foods' private infrastructure. The company is ISO/IEC 27001:2022 certified and has completed a SOC 2 Type II examination. Every customer environment is siloed. Customer data is never shared or used to train outside AI models. For organisations with the strictest data sovereignty requirements, AKA Atlas supports on-premise and air-gapped deployment: the full stack runs on the customer's own infrastructure, and nothing leaves their boundary.

Availability

AKA Studio Alchemy is available now to all AKA Studio customers. AKA Atlas is entering an invite-only beta: companies interested in early access can apply at aka-food.com or contact [email protected].

Media contact: Olga Sukhman, AKA Foods, [email protected], +1 888 301 5010.