AI agents at work: Cera moves from pilots to scale
UK home care provider Cera is deploying close to 1,000 AI agents across a 10,000-strong workforce to cut admin, speed up hiring and lift care quality amid severe staffing shortages. The agents automate routine tasks, make decisions from gathered information and act autonomously using reasoning, planning and memory.
With around 500,000 carer and nurse applications processed each year, the company is using AI to reduce bottlenecks without lowering standards. "Our AI agents remove paperwork so carers can get back to caring," says Dr Ben Maruthappu MBE, founder and CEO of Cera. "They also accelerate recruitment, ensuring more patients get better care, faster."
What's being deployed
- AI recruitment agent (Ami): conducts initial interviews, doubling recruitment volumes and easing pressure on human recruiters. Now being licensed externally.
- AI Care Coordinator Agent: halves the time spent organising last-minute cover, giving coordinators hours back per day.
- Field Care Supervisor Agent: syncs clinical data into clear summaries, cutting care review time by 85% and supporting continuous quality checks.
- Retention Agent: spots staff at risk of leaving and intervenes up to seven times faster than human teams, contributing to retention improvements of up to 22%.
- Predictive analytics: reduces falls by 20% and prevents more than half of avoidable hospitalisations.
"By automating repetitive tasks, we're enabling carers, nurses and coordinators to focus on what truly matters: care quality, health outcomes and human connection," Ben says.
Frontline impact
"We have now got time to ring clients, find out how care is going, fix issues and chase doctors and district nurses instead of spending hours of each day on admin tasks like organising cover," says Lucy Kruyer, Registered Manager at Cera Colchester in Essex. "Our AI care coordinator agent organises staff cover so we can focus on client medical and quality needs."
She adds: "We want to be able to say yes to every person that needs care. Our goal is to get people home from hospital and cared for where they want to be and the agent is helping us towards that aim."
According to the company, adult care currently faces about 110,000 vacancies, with one million new care workers needed over the next 14 years. Two million adults in England live with unmet care needs, so time saved at the front line matters.
What leaders can do now
- Start with 2-3 high-friction workflows (recruitment screening, shift cover, care reviews) and set clear SLAs and escalation rules.
- Design for safety: log decisions, enforce audit trails, and keep human-in-the-loop for risk-bearing calls.
- Pilot with measurable targets: time-to-hire, time-to-fill shifts, review cycle time, falls per 1,000 client-days, avoidable admissions.
- Integrate with scheduling, HRIS and EHR systems to avoid swivel-chair tasks.
- Train teams on new workflows, not just tools. Reward adoption and feedback.
- Establish data governance early: access controls, DPIAs, retention policies and model monitoring.
Pharma's next move: NVIDIA and Eli Lilly build a $1bn AI lab
NVIDIA and Eli Lilly are investing $1 billion over five years in a San Francisco Bay Area co-innovation lab to build AI systems for biology and chemistry, and to apply robotics and "physical AI" to drug development. The effort builds on Eli Lilly's AI supercomputing work and NVIDIA's platforms to compress discovery timelines and improve production capacity.
A core focus is foundation and frontier models trained on high-quality bio data using NVIDIA's BioNeMo platform. The goal: explore vast design spaces in silico before making a single molecule, then feed wet-lab results back into models for continuous improvement.
"AI is transforming every industry and its most profound impact will be in life sciences," says Jensen Huang, NVIDIA Founder and CEO. David Ricks, chair and CEO of Eli Lilly, adds: "By bringing together world-class talent in a start-up environment, we're creating the conditions for breakthroughs that neither company could achieve alone."
Continuous learning, from wet lab to factory floor
The partners plan a continuous learning loop connecting agent-driven wet labs with computational dry labs. Beyond discovery, they will explore multimodal models, robotics and digital twins across clinical development, manufacturing and commercial operations to strengthen supply reliability and scale production of high-demand medicines.
What hospital and life sciences leaders can apply
- Close the loop: link data generation, model training and real-world validation so systems keep getting better.
- Think beyond models: invest in workflow automation, robotics and simulation where they reduce variance and delays.
- Build cross-functional squads (clinical, data, engineering, QA) with clear ownership and decision rights.
- Co-invest with partners where infrastructure and datasets are too large to build alone.
Metrics that matter
- Home care: time-to-hire, time-to-fill shifts, care review cycle time, falls per 1,000 client-days, avoidable admissions, staff churn, client and staff satisfaction, audit findings per inspection.
- Biopharma: model-to-experiment cycle time, candidate success rates, batch release lead time, deviation rates, yield, supply fill rate.
"At Cera we are building AI that protects the human touch, rather than replacing it," Ben says. "It's about giving our staff the time, headspace and support they need to deliver exceptional care."
Helpful resources
- NVIDIA BioNeMo for domain-specific AI tooling in biology and chemistry.
- Skills for Care: Workforce state of the sector for vacancy and retention benchmarks.
If you're planning AI upskilling for clinical operations, care coordination or health administration, explore practical training by role at Complete AI Training.
Your membership also unlocks:
