☰

Aleph Farms

One process, many sites, one data path

Industry
Cultivated Meat (Cellular Agriculture)
Headquarters
Rehovot, Israel
Public information as of
March 2026

A4BEE prepared this analysis from publicly available sources. It reflects our own reading of Aleph Farms's published strategy and is not endorsed by, or produced in cooperation with, Aleph Farms. Company website

Strategic priorities

Aleph Farms holds the world's first regulatory approval for cultivated beef, granted by the Israeli Ministry of Health, and is pursuing additional approvals in Thailand, Singapore, Switzerland, the United Kingdom and the European Union. Its first commercial product, Aleph Cuts thin steaks, is sold in Israel through a B2B partnership, with regulatory filings under review in several other jurisdictions.

The company has reorganised around capital efficiency. Following a roughly 30 percent workforce reduction in 2024 and a $29 million bridge round in early 2025, it now describes itself as an asset-light producer: research, development and a pilot plant stay in Rehovot, while commercial production is run by contract development and manufacturing partners in Kemptthal (Switzerland, with The Cultured Hub), Singapore (Cell Agritech) and Thailand (a joint venture with BBGI and Fermbox Bio). The target is unit-economics-level profitability by 2028 on Platform 1.2, a single stirred-tank bioreactor process that has already reduced differentiation time by about 60 percent.

A 65,000 square foot pilot facility in Rehovot and a retrofitted manufacturing plant in Modi'in (acquired from VBL Therapeutics in 2023 for $7.1 million) sit alongside the CDMO partners. BioRaptor, an AI analytics partnership with the company of the same name, is positioned as the platform for optimising Platform 1.2 at scale. Public commitments include operational Net Zero by 2025, supply-chain Net Zero by 2030, and continued reporting under the United Nations Global Compact.

The pattern across these moves is consistent: data has to move between a Rehovot R&D hub, an Israeli brownfield plant and a small number of partner facilities operating different automation stacks, and the same data has to be assembled in formats that satisfy several different food-safety regulators at once. Whether the process is treated as a single process or as five parallel ones is largely a data-architecture decision.

Challenges we see

  • Digital Integration

    Treating distributed CDMO sites as one process

    Commercial production runs through partners in Kemptthal (The Cultured Hub), Singapore (Cell Agritech) and Thailand (BBGI / Fermbox Bio), each operating its own automation stack. The Rehovot R&D hub does not have continuous visibility into partner bioreactor parameters.

    Where product consistency depends on parameters that are confirmed by sampling after the run, the population under review is a whole batch. Streaming the same parameters as they are measured lets a deviation be flagged against the batch that is still running, regardless of where that batch is being made.

  • Operations Manufacturing

    Bringing the Modi'in brownfield plant into the same data path

    The Modi'in facility, acquired from VBL Therapeutics in 2023 for $7.1 million, was designed for biopharmaceutical production and runs legacy programmable logic controllers (PLCs), supervisory control and data acquisition (SCADA) and standalone instruments on pharmaceutical-era protocols.

    Replacing a control system that is working but unreachable usually costs more than the data it produces is worth. Connecting to it without altering its control logic and writing what is observed into the shared data path is the lighter-touch alternative.

  • Digital Operations

    Feeding BioRaptor with a complete data set

    The BioRaptor AI analytics partnership is positioned to optimise Platform 1.2, but the underlying bioprocess data currently sits across laboratory information management systems, offline sensor logs and manual spreadsheets.

    An AI optimisation model is only as good as the dataset it sees. Where continuous sensor telemetry and offline biological assays live in different systems, the model sees a partial picture and the recommendations it produces reflect that.

  • Compliance Regulatory

    Producing food-safety evidence across jurisdictions

    Aleph Farms has Israeli Ministry of Health approval and active or pending submissions in Thailand, Singapore, Switzerland, the United Kingdom and the European Union. Each regulator asks for traceability from the Lucy cell bank through manufacturing in a slightly different format.

    Where the same quality record has to be expressed in several different national formats, the work of preparing each submission is multiplied. Storing the evidence once and assembling jurisdiction-specific submissions from a single source removes most of that duplication.

  • ESG Energy

    Measuring resource intensity across a partner network

    The company has committed to operational Net Zero by 2025, supply-chain Net Zero by 2030 and annual United Nations Global Compact Communications on Progress. Resource and emissions data has to be collected across Rehovot, Modi'in and external partner facilities on a continuous basis.

    Aggregating energy, water and carbon data after the fact produces an annual figure that has to be defended. The same measurements, taken at the source and aggregated as they arrive, give a figure that can be examined during the year and signed off at the end of it.

Opportunities, by urgency and business impact

Each bubble is one opportunity, numbered to match the list below. Further right means it bites sooner; higher means a bigger effect on the business. A bigger bubble means a bigger implementation effort.

Source: A4BEE analysis of public sources

  1. Reading process signals from partner bioreactors as the run is in progress

    Bioreactor telemetry from CDMO partners in Switzerland, Singapore and Thailand lives on each partner's automation stack, and the Rehovot R&D hub sees the data only when it is shared retrospectively. Sampling-based confirmation does not flag a deviation until the run is complete.

    An agreed set of parameters, a documented protocol for streaming them, and a shared place to land them give Rehovot the same view of a partner run that the partner has, so a deviation can be acted on while the batch is still in the tank.

    • Aleph Farms corporate communications on the asset-light CDMO model, 2024–2025
    • Press coverage of the Cell Agritech Singapore partnership and The Cultured Hub Kemptthal MoU

  2. Connecting the Modi'in plant's existing equipment to the shared data path

    The Modi'in facility, acquired from VBL Therapeutics, runs PLCs, SCADA and standalone instruments on protocols designed for a biopharmaceutical site. The equipment is working but its data is not yet on the same path as Rehovot and the CDMO partners.

    Connecting to existing controllers through documented industrial protocols and writing what is observed into the shared path lets the plant's data reach Rehovot without disturbing the control logic that keeps it running.

    • Aleph Farms acquisition of VBL Therapeutics manufacturing facility, Modi'in, 2023
    • Aleph Farms production and operations public statements

  3. Unifying sensor streams and biological assay results for BioRaptor

    BioRaptor is intended to optimise Platform 1.2, but the bioprocess data it needs is split between continuous sensor streams and offline biological assay results held in separate systems, including the laboratory information management system.

    An automated ingestion layer that timestamps continuous and offline measurements against a shared batch identifier gives BioRaptor the complete view of a run that its optimisation work depends on.

    • Aleph Farms and BioRaptor AI analytics partnership announcement
    • BioRaptor's CEO, public commentary on bioprocess data volumes

  4. Generating food-safety submissions from a single evidence base

    Israeli, Thai, Singapore, Swiss, UK and EU food-safety regulators each ask for traceability documentation in a slightly different format. The current process compiles quality records manually for each jurisdiction.

    A central evidence base that holds the records once and assembles each jurisdiction's submission package on demand removes most of the duplication and makes each submission traceable back to the same underlying record.

    • Aleph Farms submissions to Thailand FDA, Singapore Food Agency, Swiss FSVO, UK FSA and EU EFSA
    • Israeli Ministry of Health cultivated beef approval announcement, 2024

  5. Automating energy, water and carbon measurement across the partner network

    Operational Net Zero by 2025 and supply-chain Net Zero by 2030 require continuous measurement of energy, water and carbon intensity across Rehovot, Modi'in and external CDMO partners. There is no automated collection layer across all sites today.

    Edge sensors and an automated aggregation layer give a per-facility and per-batch view of resource intensity that supports the United Nations Global Compact reporting cycle and the sustainability claims made to customers and investors.

    • Aleph Farms Net Zero and United Nations Global Compact commitments
    • Aleph Farms sustainability communications, 2024–2025

What we'd propose

  • Digital CDMO

    Streaming partner bioreactor data into a Rehovot data path

    A documented set of bioreactor parameters, a standard streaming protocol and a shared landing place give Rehovot real-time visibility into CDMO partner runs without disturbing partner control systems.

    • Documented parameter set

      What every partner publishes

      Define the bioreactor parameters (temperature, pH, dissolved oxygen, agitation, feed rate) that every partner site streams, and the units and metadata each value carries, so a parameter from Kemptthal can be compared with the same parameter from Singapore.

    • Streaming protocol

      How the data leaves the partner

      Specify OPC UA (Open Platform Communications Unified Architecture) or MQTT (a lightweight messaging protocol used widely in industrial telemetry) as the partner-facing transport, with role-based access that protects partner IP while making the agreed parameters visible.

    • Rehovot landing place

      Where the data arrives

      Land the streams in a documented time-series store with batch and run identifiers attached, so a Rehovot analyst can ask the same question of a partner run that they would ask of a Rehovot run.

    • A deviation surfaces against the batch that is running, not the batch that already finished.
    • The same analytical question can be asked of any site, in Rehovot or at a partner.
    • Partner sites do not have to install new control systems to participate.

    We've done this before

    See how we do Digital CDMO
  • Digital CDMO

    Connecting the Modi'in plant without replacing its control logic

    A non-invasive data path from the Modi'in facility's existing PLCs, SCADA and standalone instruments into the shared Rehovot data path, with no change to the control logic that keeps the plant running.

    • Industrial protocol integration

      Reading existing controllers

      Connect to the plant's existing PLCs and SCADA through documented industrial protocols (Modbus, Profibus, OPC DA — the older OPC specification) and read the values they already hold, without changing what the controllers do.

    • Predictive maintenance signal

      Watching equipment health

      Capture temperature, vibration and performance patterns from critical subsystems such as cooling jackets and agitation motors, and flag drift against a normal operating envelope so maintenance is scheduled before a failure.

    • Shared data path

      Modi'in data next to Rehovot data

      Write the extracted values into the same documented schema used by Rehovot and the partner sites, so Modi'in production data can be analysed alongside Rehovot research and partner production data.

    • The plant's data reaches Rehovot without replacing working equipment.
    • Equipment health is monitored continuously rather than only when something breaks.
    • Modi'in production becomes queryable in the same terms as Rehovot and partner runs.

    We've done this before

    See how we do Digital CDMO
  • Enterprise AI

    A unified data layer for the BioRaptor AI optimisation

    An automated ingestion layer that timestamps continuous sensor streams and offline biological assay results against a shared batch identifier, and delivers the harmonised dataset to the BioRaptor analytics layer in the schema it requires.

    • Continuous telemetry ingestion

      Real-time data into the lakehouse

      Pull structured telemetry from bioreactor sensors (pH, dissolved oxygen, temperature, nutrient concentration) through OPC UA into a time-series store with metadata tagging, so each measurement carries the batch, run and instrument that produced it.

    • Offline assay ingestion

      Lab results onto the same timeline

      Automate ingestion of cell density, viability and metabolite measurements from the laboratory information management system and the instruments that produce them, aligning each offline result with the continuous stream for the same run.

    • BioRaptor data delivery

      What BioRaptor receives

      Deliver the harmonised dataset to BioRaptor through a documented interface in the schema its optimisation models expect, with data-quality scoring and completeness validation so the AI sees what it has been told it is seeing.

    • BioRaptor sees a complete picture of each run rather than a partial one.
    • The time from measurement to optimisation recommendation shortens because the data path is already built.
    • Data quality issues are visible at ingestion, not discovered by the AI.

    We've done this before

    See how we do Enterprise AI
  • Digital Lab

    Generating multi-jurisdiction submissions from a single evidence base

    A central electronic batch record and quality management system that holds traceability from the Lucy cell bank through manufacturing once, and assembles each food-safety regulator's submission package from the same underlying evidence.

    • Electronic batch records

      One record per batch

      Capture each cultivation step from cell thaw to packaged product in an electronic batch record with electronic signatures, automated instrument capture and an audit trail that holds up under inspection.

    • Jurisdiction-specific assembly

      One evidence base, many submissions

      Maintain a structured map of Israeli, Thai, Singapore, Swiss, UK and EU evidence requirements, and assemble each regulator's submission package from the underlying records so the same fact is not re-keyed for each jurisdiction.

    • Cell-line genealogy

      Traceability from Lucy forward

      Track the Lucy cell bank and every derived expansion and differentiation step in a tamper-evident record, giving regulators a continuous chain of custody that is also usable in customer transparency communications.

    • Each regulator sees a complete package without manual re-keying.
    • Adding a new jurisdiction becomes a configuration change, not a documentation project.
    • Inspections are answered from the record, not reconstructed for them.

    We've done this before

    See how we do Digital Lab
  • Agents

    AI agents for multi-jurisdiction regulatory document work

    Narrow, reviewable agents that take the repetitive part of regulatory document work across Aleph Farms' active jurisdictions: drafting food-safety submission sections from source records, checking each submission package against its template before review, and identifying every record a regulatory change affects. A named reviewer signs off every output.

    • Drafting from source records

      First drafts from the evidence base

      Generate the first draft of a submission section directly from the central evidence base, so the regulatory author edits and judges rather than assembles text from the underlying records.

    • Template and completeness checking

      Gaps caught before review

      Check a draft submission against the relevant jurisdiction's template and Aleph's own checklist, returning missing or inconsistent sections before the document enters human review.

    • Change impact search

      Which records a regulatory change touches

      When a regulator updates its evidence requirements or a new jurisdiction is added, identify every record and submission section that references the change, so the scope of the update is known on day one.

    • Regulatory review queues move faster because submissions arrive complete.
    • The scope of a regulatory change is established by search rather than by recollection.
    • Each output is traceable to the source records it came from and signed off by a named reviewer.

    We've done this before

    See how we do Agents

Digital maturity: today and target

Scored out of 100 across six dimensions. The target is what Aleph Farms's own published ambition implies — not a perfect score.

Source: A4BEE analysis of public sources
Data integration 30 → 82
Continuous sensor telemetry, laboratory information management system data and offline biological assay results live in separate systems across Rehovot, Modi'in and partner facilities, and the unified path that BioRaptor depends on is ahead of the company.
Process automation 42 → 80
Platform 1.2 has reduced differentiation time by about 60 percent, but the digital process control and automated data pipelines required to repeat that performance across partner sites are still being built.
IT/OT convergence 28 → 78
Rehovot and the partner sites run different automation stacks, and the Modi'in facility's legacy controllers are still being integrated. A documented protocol set and a shared landing place for partner telemetry are the first moves.
Regulatory compliance 38 → 88
The Israeli approval demonstrates the company can satisfy a regulator, but the simultaneous submissions in Thailand, Singapore, Switzerland, the United Kingdom and the European Union require a central evidence base that is not yet in place.
Sustainability monitoring 30 → 78
Net Zero commitments are public and United Nations Global Compact reporting is in motion, but the continuous measurement of energy, water and carbon across the partner network is not yet automated.
Analytics and AI 40 → 82
The BioRaptor partnership shows the AI ambition, but the harmonised training data infrastructure that lets the optimisation models see a complete run is incomplete and is the current bottleneck.

Check this yourself

Our Service Portal has free self-assessments and market comparisons. These are the ones that line up with what we've read above — no sales call required.

Think we've read this right?

Talk to us

Related reading

This is an independent analysis prepared by A4BEE from publicly available information as of March 2026. It reflects A4BEE's own interpretation and opinion, is not affiliated with, endorsed by, or verified with Aleph Farms, and may be incomplete or inaccurate. All company names and trademarks are the property of their respective owners. To request a correction or removal, contact [email protected].