Vendor-neutral compiler for automated biology labs that turns SOPs into device-ready runs and recovers failed assays automatically.

1. A commercially announced stack now spans protocol design, device execution, and wet-lab feedback, so buyers can imagine software sitting between SOP authors and instrument runtimes.
2. Protocol reasoning performance is approaching human-expert territory, which lowers the trust barrier for constrained assay-compilation workflows.
3. Failed-step regeneration means recovery logic can be productized instead of recreated manually by scarce automation engineers after every incident.
4. The launch is framed around physical execution rather than a generic lab copilot, signaling a real market shift toward software that controls heterogeneous wet-lab work.

Catalyst. A verified July 2026 launch now shows a credible closed-loop protocol-to-device stack with wet-lab feedback and failure regeneration, making a vendor-neutral execution compiler newly believable.

The product starts as a protocol compiler that sits above a lab's existing hardware. Scientists upload an SOP or structured protocol, and the system decomposes it into reagent, timing, labware, and measurement steps before mapping each step to installed devices and operating constraints. Before execution, it simulates likely failure modes such as missing volumes, incompatible consumables, unsafe timing, or instrument contention and flags issues for human review. During the run, it ingests device logs and assay outputs, detects deviations, and proposes approved recovery or rerun instructions instead of forcing engineers to debug from scratch. Every validated run becomes reusable execution IP for the next assay transfer, customer program, or site rollout.

What's different. This is not another general "AI scientist" and not a single-vendor lab OS. The wedge is a vendor-neutral execution compiler plus failure-replay layer that works on top of a lab's existing fleet while keeping humans in the approval loop where trust matters. The moat compounds from device adapters, protocol normalization, and a growing corpus of validated recovery paths that competitors cannot easily reconstruct from static SOPs alone.

Jobs to be done

flowchart LR
  Buyer[Head of Lab Automation] --> Pain[Scientists write SOPs but engineers hand-translate device runs]
  Pain --> Product[Protocol compiler and failure replay layer]
  Product --> Outcome[Faster assay launch and fewer failed runs]

Executive takeaways

• The pain is real: labs already buy robots, but protocol-to-device translation and failed-run recovery still sit in the heads and scripts of a few automation engineers.
• The opening is between vendor-specific orchestration and higher-level informatics or cloud labs; no clear incumbent owns vendor-neutral SOP compilation plus recovery across mixed fleets.
• The best early buyers are mixed-fleet assay teams with repetitive qPCR, ELISA, or cell-based workflows, deadline-driven assay transfer, and only a small automation staff.
• Adoption will turn on trust, auditability, and implementation depth more than model novelty; human review and versioned recovery logic are table stakes.
• Technology momentum is stronger than public buying evidence, so the next step is design-partner discovery around setup time, rerun rates, and who actually controls budget.

Market definition

Vendor-neutral execution software for automated biology labs that turns scientist-authored protocols into device-ready runs, monitors deviations, and suggests recoveries across mixed workcells. The initial wedge is repetitive qPCR, ELISA, and cell-viability workflows inside synthetic biology CROs, biofoundries, and translational core labs.

Customer and buyer

Primary users are lab automation engineers and assay-development leads who own workcell setup, debugging, and transfer. The economic buyer is usually a head of lab automation, platform-operations leader, or core-facility director who feels the cost of slow launches and reruns across multiple instruments.

Buying triggers

• A new assay family, second site, or new workcell forces the team to translate one protocol across multiple instruments and software layers. [12][13][14][18]
• Throughput goals rise faster than staff capacity, so teams need more walk-away time and fewer manual handoffs between liquid handlers, incubators, washers, and readers. [14][15][29][27]
• Labs running translational or validated workflows need auditability and electronic-record controls when automation logic changes or runs deviate. [17][21][30][31][32][33]

Willingness to pay

Willingness to pay is credible because buyers already fund expensive adjacent layers: orchestration software, validated informatics, cloud-lab access, and workflow scheduling. A compiler priced around a workcell budget line can be justified if it meaningfully cuts engineer setup time and reruns without forcing hardware replacement. [40][26][27][16][20]

Category dynamics

Growth signal 8.4% CAGR

Validation signals

• ProtoPilot is direct evidence that protocol-to-device generation with wet-lab feedback and failure regeneration is now a credible product surface.
• NIST and NSF are treating autonomous and programmable laboratories as a serious infrastructure trend rather than a novelty.
• CoreMarketplace and biofoundry examples show a real population of automation-heavy labs that could act as design partners.
• Benchling, Biosero, HighRes, and other orchestration launches show that buyers already understand the adjacent category, even if the exact compiler wedge is still open.

Regulatory & technical constraints

• Execution software that touches translational or validated workflows must support audit trails, controlled changes, and durable electronic records.
• Open standards help, but real deployments still require adapter work for heterogeneous instruments, schedulers, and data outputs.
• Assay automation spans more than liquid transfer; real workflows tie together handlers, incubators, washers, readers, and analysis systems.
• Human approval remains a practical control point for recovery logic and reruns even when closed-loop generation improves.

Competition is fragmented rather than empty. Biosero, Thermo/Tecan, and other automation vendors own device orchestration; Benchling is pushing hardware-agnostic automation and scientific records; Emerald Cloud Lab sells the full remote-lab substitute; and Opentrons lowers the cost of standardizing on one stack. The new startup must therefore win on vendor-neutral compilation and recovery, not on generic automation rhetoric.

Why incumbents do not win by default

• Vendor-specific automation software. Thermo/Tecan-class workflow software is strong at scheduling and controlling integrated systems, but it does not win by default when a lab needs neutral translation across mixed fleets and human-reviewable recovery logic.
• Scientific informatics platforms. Benchling-class platforms increasingly connect instruments and records, but they still frame the problem as record-centric automation and analytics rather than as SOP-to-device compilation plus runtime recovery.
• Cloud labs. Emerald Cloud Lab proves demand for remote automated execution, but it is a substitute built around the provider’s operated facility rather than a control layer for the customer’s installed fleet.
• Single-stack hardware ecosystems. Opentrons and similar platforms make protocol authoring far easier when a lab standardizes on one stack, yet mixed-fleet labs still need cross-device abstractions, adapters, and exception handling.
• In-house scripting and services. Manual integrations remain common because labs can always fall back to engineers, consultants, and workflow-by-workflow setup, but the pain sources show why that approach scales poorly.

Mixed-fleet biology labs already own liquid handlers, incubators, and plate readers, but translating scientist-authored SOPs into device-ready runs and recovering failed assays still depends on a few automation engineers. We start with 50-300 person synthetic biology CROs and translational core labs because new customer assays, new workcells, and second-site transfers force the same protocol across heterogeneous instruments under deadline. The product is a vendor-neutral execution compiler that ingests an SOP, maps it to supported devices and labware, simulates likely failure points, and outputs a human-reviewable run plan plus approved recovery steps rather than replacing the lab's installed orchestration stack. The first proof point is narrow and falsifiable: cut time from SOP handoff to first validated automated run on one repetitive assay family, then reduce failed-run recovery time on that same workflow. Research-sized demand is meaningful but not infinite: TAM is estimated at $412.5M, beachhead SAM at $148.5M, and the year-3 SOM at roughly $5.0M under a high-touch design-partner motion. The company wins only if device adapters, validated protocol mappings, and approved recovery histories compound into execution IP that vendor-native tools and internal scripts cannot reproduce across mixed fleets. The biggest disconfirming risk is not model novelty but whether the pain is frequent and budget-owning enough to support software rather than custom integration services. Research did not independently confirm customer adoption, production error rates, or broad instrument coverage beyond one credible launch, so this plan deliberately keeps scope narrow, human-in-the-loop, and pre-seed sized.

Problem

• Labs with mixed robotic workcells still spend days translating each new qPCR, ELISA, or cell-viability SOP into vendor-specific scripts, making 2-5 automation engineers the bottleneck for assay launch, transfer, and troubleshooting.
• When a run fails, device logs, labware assumptions, and recovery logic are fragmented across tools and individual engineers, causing reruns, sample loss, and weak auditability in translational workflows.

Solution

• Compile a scientist-authored SOP into a supported device graph, labware plan, and execution sequence for one assay family, with simulation checks for timing, consumables, volume, and instrument contention before the run starts.
• Ingest device logs and assay outputs during execution to surface deviations, propose human-reviewable recovery paths, and store a versioned execution record that can be reused for future assay transfer and site rollout.

Why we win

• The category gap is specific: incumbent orchestration tools control devices, and informatics platforms manage records, but no named competitor clearly owns vendor-neutral SOP-to-device compilation plus runtime recovery across mixed fleets.
• Every validated compile adds protocol-to-device mappings, constraint data, and approved recovery histories that become harder for OEM software or internal scripts to replicate across heterogeneous workcells.
• The beachhead pain is tied to hard delivery deadlines and scarce labor, so a narrow workflow product can prove ROI faster than a broad "AI scientist" platform.

Milestones

flowchart LR
  Wedge[Mixed-fleet assay beachhead] --> MVP[Protocol compiler MVP]
  MVP --> Proof[Setup-time and recovery proof]
  Proof --> Expansion[Recovery library, compliance, multi-site rollout]

Founding team

Experiment roadmap

Risk assessment

What must be true

• Target beachhead labs spend 40+ automation-engineer hours to move a supported assay from SOP handoff to first validated automated run.
• At least 25% of major reruns or launch delays in the beachhead stem from execution and integration issues the product can address rather than assay science alone.
• One narrow adapter set covers 60%+ of relevant qPCR, ELISA, and cell-viability volume at the first design partners without hardware replacement.
• The economic buyer will fund a workcell-based software line item after a paid pilot instead of treating the problem as project-based services.
• At least half of paid pilots convert to annual production contracts within 6 months because setup-time and recovery metrics improve enough to justify rollout.

Open diligence questions

• What baseline SOP-to-first-run setup times and rerun rates did named target labs share, and how were they measured?
• Which instrument families dominate the first 10 target accounts, and how much adapter overlap is actually reusable across them?
• Who signs the first contract and from which budget line is the spend reallocated?
• How much of recovery guidance can be standardized from telemetry versus encoded manually by services staff?
• What prevents Benchling, Biosero, or OEM workflow software from adding enough compiler and recovery capability to erase the wedge?

Model sanity

• Revenue engine. The base case reaches $3.6M of Y3 revenue by turning three paid design partners into recurring customers, exiting Y2 with 8 paying labs, and expanding the base toward 15 labs by Q4Y3.
• Must go right. Adapter reuse has to keep onboarding product-like enough for gross margin to climb from the pilot phase into the low-70s by Y3.
• Model breaks if. If pilot-to-production conversion slips a quarter and gross margin stalls below 68%, the downside case runs through cash before the next round.
• Next-round proof. The next financing story is 2-3 converted production contracts with measured ROI by month 18 and roughly 8 paying labs by Q4Y2.

Scenarios

Sensitivity

flowchart LR
  Accounts["Target accounts"] --> Pilots["Paid pilots"]
  Pilots --> Contracts["Production customers"]
  Contracts --> Workcells["Covered workcells"]
  Workcells --> Subscription["Subscription revenue"]
  Pilots --> Onboarding["Pilot and onboarding fees"]
  Onboarding --> Revenue["Total revenue"]
  Subscription --> Revenue
  Revenue --> GrossProfit["Gross profit"]
  GrossProfit --> Cash["Ending cash"]

Flags: The model leans heavily on three Y1 design partners converting on schedule, so one missed conversion would delay the Q4Y2 proof point materially. · Gross margin only reaches the BP target if adapter reuse is real by late Y2; otherwise delivery looks more like integration services than software. · Q4Y3 exit ARR is near the researched $5.0M SOM, but recognized Y3 revenue is lower because the customer adds happen throughout the year. · Rule-of-40 optics look unusually strong because the business is coming off a small base; renewal quality and onboarding reuse matter more than the headline ratio. · Cash roll-forward uses EBITDA as a proxy for cash, so procurement delays or larger receivables swings could compress the modeled cash floor.

• Device-integration sprawl. Supporting heterogeneous lab hardware could overwhelm the team before product-market fit is proven. Mitigation: Start with one assay family and a narrow adapter set around the most common workcell combinations in design-partner labs.
• Validation and trust barrier. Labs may hesitate to let software influence wet-lab execution without proof that compiled runs are safe and reproducible. Mitigation: Launch with human-in-the-loop review, simulation checks, run diffing, and auditable approvals before expanding autonomy.
• Thin independent evidence. The current market signal rests on a single verified announcement, so real buyer urgency or adoption could be overstated. Mitigation: Target customers already staffing around the pain, run paid pilots against setup-time and failed-run metrics, and kill the wedge quickly if budget authority is weak.