Turning Bioreactor Data into Competitive Advantage for a Next-Generation Cultivated Meat SaaS Platform
How Xcelore designed and engineered Pythag — a real-time monitoring and analytics platform with IIoT integration across 75 bioreactors, conversational AI, and predictive intelligence for cultivated meat production.
DATA ANALYSIS TIME
- 40% reduction
INSIGHT ACCESS SPEED
+50% faster
CROSS-TEAM COLLABORATION
+30% improvment
REPORTING EFFICIENCY
+70% increase
Why Pythag Needed to Exist
A USA-based SaaS provider serving the cultivated meat industry needed their platform to evolve from a data repository into a fully instrumented intelligence system.
Cultivated meat production is an extraordinarily data-intensive discipline. Every bioreactor run generates continuous streams of sensor readings — pH, dissolved oxygen, CO₂, cell density, temperature — that must be captured, correlated, and acted on in near-real time. The client’s customers were navigating this complexity with fragmented tools, manual processes, and no unified view of their operations.
The client is a USA-based SaaS provider whose platform serves bioprocess teams at multiple production facilities across North America. With customer growth accelerating and production sites expanding, they approached Xcelore to design and build Pythag — a cloud-based monitoring and analytics platform powered by AWS, IIoT integration, and real-time data visualisation.
The brief went far beyond dashboards. It demanded an end-to-end rethinking of how bioreactor data should flow, be stored, be queried, and be acted upon — from the OPC-UA protocol layer through to a conversational AI agent that lets researchers ask questions about their data in plain language.
CLIENT
USA-based SaaS Provider
SECTOR
AgriTech/BioTech SaaS
PLATFORM
Cloud • IIoT • Web
BIOREACTOR SCALE
Up to 75 units monitored simultaneously
DELIVERED BY
Xcelore — AI & Data Engineering
Five Gaps No Existing Tool Could Close
Before Pythag, the client’s customers were operating with fragmented tools, manual workflows, and zero unified visibility across their bioreactor estate.
01
No Real-Time Bioreactor Tracking
There was no centralised system for monitoring bioreactor or batch performance. The OPC-UA communication protocol had to be built from the ground up — without it, real-time visibility and scalable production were simply not achievable.
02
Siloed Process Data
Critical parameters — pH, O₂, CO₂, cell density — were fragmented across separate systems with no unified analysis layer. This made it difficult to generate actionable insights and significantly slowed decision-making across research and production teams.
03
No Batch Management Tools
Standard off-the-shelf tools could not support SCADA-like monitoring, batch lifecycle management, or seed train tracking — creating operational inconsistencies and placing an artificial ceiling on cultivation efficiency.
04
Hidden Resource Costs
Teams had no visibility into energy consumption, water usage, or contamination rates. This blind spot drove up production costs and made sustainability reporting — increasingly expected by investors and regulators — impossible.
05
No Predictive Intelligence Layer
The existing setup generated no predictions, no pattern recognition, and no recommendations from the wealth of data produced by each run. Researchers needed a conversational AI agent that could surface insights from sensitive bioreactor data without requiring specialist data science skills.
06
Scaling Bottleneck
Without a real-time data infrastructure, scaling cultivated meat production across multiple facilities was simply not feasible. The platform needed to be architected for production-grade reliability from day one.
What Xcelore Set Out to Achieve
A clear set of platform objectives that guided every design and engineering decision across the project.
IIoT Integration at Protocol Level
Build OPC-UA connectivity from scratch for reliable, real-time bioreactor data exchange.
Centralised Data Warehouse
Aggregate all bioprocess data securely with scalable storage and reliable cross-team access.
Conversational AI Agent
Enable natural-language querying of sensitive bioreactor data with a dual chat/build interface.
Production-Grade Reliability
Architect for scale from day one — multi-facility, high-frequency sensor data without failure.
Real-Time Dashboards & Visualisation
Deliver adaptive, anomaly-highlighting dashboards that surface actionable insights instantly.
Sustainability Metrics & Reporting
Track energy, water usage, and environmental impact data across all active cell species.
Predictive Intelligence
Surface patterns, flag anomalies proactively, and generate recommendations from historical run data.
Embedded Training & Onboarding
Guided video onboarding for device setup, seed train configuration, and platform adoption.
IIoT Integration at Protocol Level
Build OPC-UA connectivity from scratch for reliable, real-time bioreactor data exchange.
Centralised Data Warehouse
Aggregate all bioprocess data securely with scalable storage and reliable cross-team access.
Conversational AI Agent
Enable natural-language querying of sensitive bioreactor data with a dual chat/build interface.
Production-Grade Reliability
Architect for scale from day one — multi-facility, high-frequency sensor data without failure.
Real-Time Dashboards & Visualisation
Deliver adaptive, anomaly-highlighting dashboards that surface actionable insights instantly.
Sustainability Metrics & Reporting
Track energy, water usage, and environmental impact data across all active cell species.
Predictive Intelligence
Surface patterns, flag anomalies proactively, and generate recommendations from historical run data.
Embedded Training & Onboarding
Guided video onboarding for device setup, seed train configuration, and platform adoption.
Five Capability Pillars of the Pythag Platform
Xcelore engineered Pythag as a modular, microservices-based platform — five interconnected solution areas, each directly resolving one of the identified challenges.
Architected for Performance, Reliability & Scale
Every technology layer was chosen for the specific demands of high-frequency bioreactor sensor data at production scale — with observability, cost efficiency, and security built in from the start.
Frontend
React.js / Next.js (SPA) — Highcharts for live data visualisation and real-time charting.
Backend
Java-based microservices architecture — modular, independently deployable services per domain.
Databases
Infrastructure
AWS EKS · AWS Lambda · Terraform (IaC) — scalable, reproducible, and managed cloud deployment.
CI/CD & Quality
GitLab · Maven · SonarQube (code quality) · Snyk (security scanning throughout the pipeline).
Observability
New Relic (platform observability) · Ansible (configuration management) · Spot.io (cloud cost optimisation).
Measurable Results, Traceable to Every Challenge
Every outcome maps directly to one of the five challenges Pythag was designed to solve — demonstrating end-to-end resolution of the client’s core operational problems.
−40%
Reduction in Data Analysis Time
Researchers eliminated manual data searches entirely, freeing capacity for higher-value experimentation. Directly resolves the absence of centralised process data analysis.
+50%
Faster Access to Process Insights
Teams now extract critical process and performance information in half the time — boosting decision-making confidence and accelerating the pace of production iteration.
-40%
Reduction in Data Analysis Time
Researchers eliminated manual data searches entirely, freeing capacity for higher-value experimentation. Directly resolves the absence of centralised process data analysis.
+50%
Faster Access to Process Insights
Teams now extract critical process and performance information in half the time — boosting decision-making confidence and accelerating the pace of production iteration.
+30%
Improvement in Cross-Team Collaboration
Centralised data access and shared dashboards aligned research, production, and management teams — eliminating the silos that had previously slowed decisions across the organisation.
+70%
Increase in Reporting Efficiency
Automated report compilation dramatically reduced the time needed to produce comprehensive research and production reports — directly resolving the resource-visibility and sustainability reporting gaps.
+30%
Improvement in Cross-Team Collaboration
Centralised data access and shared dashboards aligned research, production, and management teams — eliminating the silos that had previously slowed decisions across the organisation.
+70%
Increase in Reporting Efficiency
Automated report compilation dramatically reduced the time needed to produce comprehensive research and production reports — directly resolving the resource-visibility and sustainability reporting gaps.
Frequently Asked Questions
These are the most commonly asked question about Xcelore.
1. What is Pythag and who is it for?
▾Pythag is a cloud-based monitoring and analytics platform for cultivated meat production, developed by Xcelore. It is designed for bioprocess research and production teams who need real-time visibility across their bioreactor estate, AI-driven process insights, conversational data querying, and sustainability reporting — all in a single unified system.
2. What technology stack was used to build Pythag?
▾Pythag was built with React.js and Next.js for the frontend, Highcharts for live data visualisation, Java microservices on the backend, PostgreSQL and InfluxDB for databases, Redis for caching, AWS EKS and Lambda for infrastructure, Terraform for infrastructure as code, and New Relic and Ansible for observability and configuration management.
3. How many bioreactors can Pythag monitor at once?
▾Pythag supports simultaneous real-time monitoring of up to 75 bioreactor units, with side-by-side comparison views and remote access from any location via OPC-UA connectivity built from the protocol layer up.
4. What were the measurable outcomes after launching Pythag?
▾Pythag delivered a 40% reduction in data analysis time, 50% faster access to process insights, a 30% improvement in cross-team collaboration, and a 70% increase in reporting efficiency through automated report compilation — with every outcome traceable to one of the five original challenges.
5. Can Xcelore build a similar AI or IIoT platform for my business?
▾Yes. Xcelore specialises in AI, IIoT, and data engineering for startups and growth-stage SaaS companies. We have end-to-end capability across cloud infrastructure, real-time data systems, conversational AI, and analytics platforms. Get in touch at contact@xcelore.com or visit xcelore.com/contact.
Ready to Build
Your Next Intelligent Platform?
Xcelore partners with startups and growth-stage companies to design and engineer AI and IIoT products that turn complex data into competitive advantage.
