ReactWise

What do they actually do

ReactWise provides a browser-based, no‑code platform that helps chemists choose better reaction and process conditions with fewer experiments. Users upload prior results or define the parameters to explore, the software builds a predictive model, recommends the next experiments to run, and surfaces visualizations and parameter importance. It includes pre‑built chemical descriptor libraries and proprietary reactivity models so users don’t have to encode every molecule from scratch (product overview, predict, insights).

The platform can connect to lab automation to run closed‑loop optimization: it sends instructions to instruments (e.g., flow reactors, analytics) and ingests results to update recommendations automatically. ReactWise also operates a small robotic wet lab to generate training data and to demo closed‑loop experiments when customers need it (automate, funding/news).

Today it’s used by bench chemists and lab leads in pharma, biotech and specialty chemicals. Named customers/partners include OnDemand Pharmaceuticals, Pharmaron, Vapourtec and academic groups, and the company reported running 12 pilots with large pharma as of March 2025 (YC profile, site, news, TechCrunch).

Who are their target customer(s)

• Bench chemist / lab experimentalist (pharma/biotech): Spends many iterative runs guessing which variables to try next; loses time re‑running failed conditions and translating messy notes into clear next steps.
• Process development lead / R&D manager: Needs reproducible, scalable recipes on tighter timelines but faces long optimization cycles, high reagent/analytical costs, and uncertainty when handing results to scale‑up teams.
• Lab automation / operations lead (robotics & instruments): Must integrate diverse instruments and run closed‑loop experiments, but deals with brittle one‑off integrations and unreliable handoffs between software and hardware.
• CMO/CRO or specialty‑chemicals R&D director: Has to deliver optimizations quickly with limited high‑quality cross‑project data; failed batches and impurities increase cost and delay customer projects.
• Cheminformatics or pharma data‑science lead: Needs clean, well‑annotated data and generalizable models but faces sparse/inconsistent lab data and time‑consuming descriptor engineering.

How would they acquire their first 10, 50, and 100 customers

• First 10: Target process‑development teams at a few large pharmas and select CRO/CMO partners; run high‑signal pilots using ReactWise’s robotic lab to deliver an optimized reaction and written recipe, then turn those into public case studies and references (news, TechCrunch).
• First 50: Package a standardized “pilot‑to‑production” offer (data ingest, model build, on‑site/remote closed‑loop demo, success metric, exportable recipe) with risk‑sharing pricing; use early references for targeted outreach to mid‑sized pharma, specialty chemicals, and CMOs, with a sales‑engineer to handle fast evaluations/integrations (solutions, YC profile).
• First 100: Scale via two tracks: hire enterprise sales/CS for conversion, compliance, and onboarding; and open channel partnerships with automation vendors and CROs, productizing common hardware integrations and offering a lab‑as‑a‑service option for non‑automated customers (TechCrunch, partners/news).

What is the rough total addressable market

Top-down context:

Global pharma R&D is about $289B (2024), with roughly 40–45% in preclinical/discovery; the chemistry/process slice of that preclinical spend supports a practical core TAM in the low‑to‑mid tens of billions (Statista, CBO).

Bottom-up calculation:

Estimate chemistry/process work as 10–25% of preclinical (~$12–31B), then conservatively add a small share of adjacent lab‑automation and chemical‑software budgets (~$3–5B) for a near‑term TAM ≈ $15–18B; with broader adoption across software/automation adjacencies, the opportunity ranges roughly $25–45B (Grand View Research, PR Newswire, MarketsandMarkets).

Assumptions:

• Chemistry and process‑development represent ~10–25% of preclinical/discovery spending globally.
• Customers will reallocate a small portion of lab‑automation and chemical‑software budgets to optimization software without double counting.
• Small‑molecule chemistry remains a material share of R&D activity, sustaining enterprise‑grade software spend.

Who are some of their notable competitors

• Kebotix: Self‑driving discovery stack coupling ML models with robotics for closed‑loop experiments; overlaps on autonomous optimization but is more materials‑focused and vertically integrated around its own robotic workflows (overview).
• DeepMatter: Captures, standardizes, and serves reaction data via APIs and tools; competes on the data/model layer for reaction prediction and reproducibility rather than instrument orchestration (profile).
• Synthace (Antha): No‑code platform to design and automate lab workflows and translate protocols to automation; overlaps on workflow orchestration but is broader and less focused on chemistry‑specific predictive reactivity models.
• Arctoris: Robotic wet‑lab service and platform executing experiments remotely to deliver AI‑ready, reproducible data; competes with ReactWise’s wet‑lab + closed‑loop demo capability (platform).
• Uncountable: Enterprise R&D data platform with DOE and predictive tools for chemicals/materials; overlaps on experiment suggestion and data management but positioned as broader lab informatics rather than a chemistry‑specific co‑pilot tied to a robotic lab (chemicals).