Revvity unveils Signals Xynthetica to close the model-experiment gap in discovery
Revvity has introduced Signals Xynthetica, an AI-augmented design platform built to speed molecular and materials discovery. Delivered as a Models-as-a-Service (MaaS) offering within the Revvity Signals ecosystem, it connects in-silico design with real experimental data under a governed, continuously learning loop. Pre-registration is open, with early access planned for the first half of 2026.
The goal is simple: bring de novo generation, property prediction, and multi-objective optimization into one environment that scientists can actually use in day-to-day work. By tying model outputs directly to lab results, teams can iterate faster and trust what the models suggest.
What Signals Xynthetica does
- De novo design: Generate candidate molecules and materials with tunable constraints based on project goals.
- Property prediction: Estimate key physicochemical, biological, or materials properties before investing in experiments.
- Multi-objective optimization: Balance potency, selectivity, ADME, manufacturability, cost, and more within a single optimization pass.
- Closed-loop learning: Benchmark model outputs against experimental results, then update models as new data arrives.
- Governance: Keep traceability on model versions, data lineage, and decision history to meet scientific and regulatory standards.
Why this matters for R&D teams
Most organizations still run strong models in isolation from the lab. That gap burns budget and time. Xynthetica's pitch is to unify design, prediction, and validation so you can move from "interesting idea" to "tested result" with fewer handoffs and fewer dead ends.
The practical win is workflow fidelity: predictions are measured against real outcomes, uncertainty is visible, and models get better with use. That keeps AI useful, not ornamental.
Signals ecosystem: the broader context
Xynthetica builds on Revvity's multi-year push to make AI native to scientific workflows:
- Signals One (April 2025): A unified data platform spanning design through decision-making, making research data AI-ready with features like semantic search and automated protocol suggestions while protecting IP.
- Living Image Synergy AI (early 2025): Multimodal imaging analysis that automates tedious steps and improves reproducibility across imaging modalities.
- Transcribe AI (late 2024): OCR for clinical labs to clean up data entry and reduce errors.
- ACD/Labs acquisition: Adds analytical and molecular design depth, strengthening data generation, interpretation, and predictive modeling inside Signals.
Access and timing
Pre-registration has started. Early access is slated for H1 2026. If you're evaluating, expect enterprise-grade deployment patterns (MaaS), integration hooks into existing Signals tools, and governance features that matter in regulated settings.
Impact on RVTY shares
Following the Dec. 16 announcement, the stock traded roughly flat through yesterday. Over the past six months, RVTY is up 1.4% versus 6.2% for its industry, while the S&P 500 gained 15.3%. Longer term, embedding AI tightly with empirical workflows could expand Revvity's role in discovery, especially as demand grows for platforms that connect data infrastructure, modeling, and lab execution.
Broader med-tech comps and recent results
- Medpace Holdings (MEDP): Zacks Rank #2 (Buy). Q2 2025 EPS of $3.10 beat by 3.3%; revenue of $603.3M beat by 11.5%. Long-term estimated growth rate: 11.4%. Average earnings surprise across the last four quarters: 13.9%.
- CareCloud (CCLD): Zacks Rank #2. Q3 2025 adjusted EPS of $0.10 beat by 25%; revenue of $31M beat by 8.3%. Estimated growth rate for 2025: 20%.
- KORU Medical Systems (KRMD): Zacks Rank #2. Q3 2025 adjusted loss per share of $0.02, narrower than expected by 33.3%; revenue of $10M beat by 7.1%. Estimated growth rate for 2026: 83.3%.
What R&D leaders can do now
- Make data model-ready: Standardize assay metadata, units, and ontologies; enforce data quality checks at ingestion.
- Define objective functions: Agree on weighted criteria (e.g., potency vs. solubility vs. cost) before optimization runs.
- Wire the feedback loop: Automate the flow from model proposals to experiment design, results capture, and retraining.
- Set governance early: Version models and datasets, log decisions, and make uncertainty and rationale visible to reviewers.
- Skill up the team: Upskill scientists on prompt patterns, model limits, and validation methods; train data stewards to keep pipelines clean.
Bottom line
Signals Xynthetica is built to reduce the friction between AI-generated candidates and what the lab proves out. If your roadmap includes AI-guided design, prioritize data readiness, objective clarity, and a closed-loop workflow. That's where the time savings and better hit quality show up.
Looking to upskill your team on practical AI for scientific work? Browse courses by job role.
Your membership also unlocks:
