Adam Sroka

Data & AI Director

Hypercube Consulting

Chief Executive Officer

Director

The Data Lab - Innovation Centre

The Innovation Advisory Board is responsible for advising on innovation and collaborative project activity for The Data Lab – Scotland’s Innovation Centre for Data and AI.

Innovation Advisory Board Member

Origami

Head of Machine Learning Engineering

Incremental Group

Senior Data Scientist

Sure Thing! Insurance

Data Scientist

Clear Returns

Thales

Research Engineer

I build data and AI platforms for the energy sector | Director at Hypercube Consulting

Dr Adam Sroka is a Co-founder and CEO at Hypercube Consulting, a firm that helps organisations unlock value from data by delivering enterprise-scale solutions and building high-performing data and analytics teams from the ground up. With over 10 years of experience in data science, machine learning, and artificial intelligence, Adam is a recognised expert and speaker in his field.

Adam's mission is to bridge the gap between technology and business, and to enable data-driven innovation and transformation across various industries. He shares his insights and best practices through his blog, podcast, and tech community events, where he showcases his approaches and systems for effective communication, engineering discipline, and experimentation. Adam is passionate about helping organisations and data professionals maximise their potential and impact with data and AI.

MiiHealth AI

- Built Mediiflow to overcome the challenges of existing systems for Voice AI and IVR, specifically for complex clinical triaging and guideline-focused symptom-first patient intake protocols.
- Outcomes from Live Pilot from Mayo Clinic Arizona so far: ~8.2 min saved every visit, 200+ patient intakes successfully executed so far for the cardiology department, 92% CSAT score!

Where we win (vs incumbents):

- Against those who optimise forms/RPA - we deliver conversational voice and clinician-centric build (fast time-to-value). 
- Against those who are telephony-centric -  our multi-modal (kiosk→phone→avatar) keeps a single protocol across touch-points.

My role:

- Maintaining/Improving Mediiflow (engine behind DAINA patient-intake agents); led data/AI/voice infra and Agent Builder productization.
- Running AI R&D → demo-safe deployments with clinics/hospice teams; 
- Driving privacy-by-design architecture (per-department isolation, in-tenant LLM/STT/TTS)

Why we matter:

- Ops leaders: free front-desk minutes; cut no-shows & intake friction; improve throughput without headcount.
- CIOs/CISOs: single-tenant stamps reduce review risk; auditability and in-boundary AI by default.
- For providers: Walk into patient visits with HPI summarised and symptom red flags highlighted, ready for EHR entry.
- For PCNs: reduce burden to keep queues safer and on the flow by letting DAINA do the intake autonomously in-person via kiosks, computers, tablets or via phones (scheduled ahead of time or immediately!)

Keywords (discovery):

Patient intake, multi-modal voice, WebRTC, Agent Builder, HIPAA, deployment stamp, pre-op, reminders, triage, FHIR, Epic/Cerner, departmental pilots.

Co-Founder

MiiCare

- Onward Care @ Buckinghamshire (with Sodexo): avoidable readmissions cohort (high social need) → bed-days reduced by 77%; 83% ROI for elective cases, and 35% ROI for non-elective cases; £287k savings so far; Freed capacity for winter pressures.
- Falls prevention (BLMK ICS / home & care-at-home): falls −76% over 12–18 months; 6.2× ROI; acute cost avoidance documented at £158k tangible savings and £1.25m intangible long term savings; older adults stayed at home longer by 497 days.
- Chronic care (InnovateUK/SBRI with DDM Health): deployed Monica’s recommendation engine with Low Carb Program for older adults living alone (T1/T2) → measured gains in function, sleep, mobility; program delivered end-to-end.

Evidence & IP

- Patent: Mobility Analysis using Acoustic Gait Detection (GB2607561).
- Publications: 7 peer-reviewed papers on acoustic gait & voice biomarkers predicting fall/cognitive risk.
- Longitude Prize Dementia (AgeUK & InnovateUK): Phase-1 participant; co-authored successful bid (multi-year program funding secured).

What I built / led
- PHI stack for the home: multi-signal data pipeline (ambient, wearables, conversational - MiiVoice) powering proactive alerts and next-best actions for carers and clinicians.
- Proactive CAI (Monica) + MiiGait: privacy-safe nudges and passive mobility risk—designed to act before incidents, not just report them.
- Governance: GDPR/IG-aligned data management, clinical safety processes, de-identified analytics, and commissioner-ready dashboards for D2A, falls, hydration

Team & partnerships: led DS/AI from R&D → IP → deployment; structured collaborations with NHS Trusts/ICBs, Sodexo, AgeUK, DDM Health; authored/secured UK grants.

Why it matters (for ICPs)

- ICBs & Trusts: free beds faster (bed-days ↓), reduce DTOC risk, protect elective recovery.
- Local authorities & providers: prevent falls, right-size visits, lift staff capacity without adding headcount.

DM me with any questions you may have!

Interim Chief Data Officer

-  Delivered a 2-year long InnovateUK project (ADAPTIVE) as the AI lead from MiiCare which demonstrated how the acoustic gait analysis via the MiiCube can be used for early detection and improved clinical outcomes of cognitive decline in older adults. Partners: Harmonia Village, Graham Care, Avante Care, University of Kent (Medway)
- Wrote the Technical Documentations which lead to MiiCare winning 3 grants from the UK Government (Smart Grant, InnovateUK & SBRI) with a combined funding of over £1 million
- Helped MiiCare file 2 patents for MiiCube as a Acoustic Gait Sensing Technology
- Designed and implemented MiiCare's Data Infrastructure which enables MiiCare's Smart Insights Engine, MiiCortex
- Optimised cost of MiiCare's Data Infrastructure by 87% with operating expenses less than £500 per month
- Grew the Data Science Team by adding 3 clinical analysts and 1 Data Engineer
- Lead the Clinical Data Science team to create novel algorithms which relies on minimal number of sensors (smart devices are optional sources) to extrapolate sleep quality and physical wellness of older adults. 
- Lead the Clinical Data Science team to deliver 4 new types of smart insights: MiiSleep Score, MiiWellness Score, Lower UTI onset detection and Fall Events in bathrooms/washrooms.
- Mentored 2 interns which are set to join the Data Science team as FTEs at the end of Q4 2022

Data Science Lead

- Launched a Free Offering (named "MiiCough") of the RTI detector to the general public for Dataset Crowdsourcing and Machine Learning Research on 12th March 2021
- Deployed a prototype of an Respiratory Tract Illness (RTI) Detector using Acoustic Machine Learning using a 3-Node Kubernetes Cluster on Azure Kubernetes, using Azure DevOps
- Lead and Core Data Engineer to start building the first iteration of the Behavior Anomaly Detection and Clinical Data Analytics Platform; currently using Azure Gen2 Datalake Storage, Azure Databricks and MODE Analytics
- Exploring possible clinical insights to be extracted from existing customer base and pilot projects
- Second Author for scientific paper in the Gait Analysis Research; publication currently pending in IEEE 
- Conducted Machine Learning research for novel application of Acoustic Gait Analysis for understanding elderly health
- Attended 2 Microsoft Virtual Hacks (on private invitation) to learn about Azure Data Factory, Azure Databricks, Azure Functions, Azure Kubernetes and Azure DevOps

- Extracted, cleaned and obtained insights from IOT Sensor data
- Conducted an extensive 8-month research on Assisstive Technology for the Elderly under the supervision of the CEO
- Co-wrote literature review on the research done with the CEO

Data Science Intern

TeamCognito

- Won “Most Innovative Idea” award for “Citizen’s Voice” Challenge hosted by the Indian Government and NASSCOM Centre of Excellence (Rural Upliftment for addressing complaints using Speech Translation)
- Public Speaking Invite at Google Developers Club: Representative Complexity for Neural Networks

Machine Learning Associate

Apnim

Created an ensemble-based 3-step Deep Learning model pipeline from scratch for Hand-written Signature Verification for Legal Documents and Banking Kiosks

Machine Learning Intern

Interim CDO | Digital Health | AI for Social Good

Voice assistants that act first with a health and wellbeing mandate — not toys, not chat. I build proactive, privacy-safe CAI strategy + infrastructure + systems that turns home data into next-best actions for older adults and busy clinical teams.

Receipts:
1. In the UK: 77% fewer hospital bed-days post-discharge; 83% ROI in high elective D2A cases; 76% fewer falls in high-risk clients; 6.2× ROI for ICS clients,  
2. In the US: Created Mediiflow, the world's first programming language + runtime for medical CAI; DAINA (built using Mediiflow) stitches the patient voice journey (kiosk → phone → avatar) on a single-tenant “deployment stamp” per department; live pilots with Mayo Clinic shows 8.2 mins saved in every intake, 92% patients loved DAINA, 300+ patient intake processed so far → double-digit ROI

What I do: 
Architect PHI pipelines (75+ in-home signals), design proactive CAI, publish on acoustic gait/voice biomarkers/new generation proactive Voice Assistants applied to healthcare pathways , and ship commissioner-safe dashboards teams actually use.

Best fit: 
NHS ICB/Home-First, councils, care providers; US service-line leaders who need capacity now (peri-op, intake, rev-cycle). 

Keywords: 
PHI, proactive CAI, discharge-to-assess, falls prevention, bed-day reduction, Family OS, acoustic gait, voice biomarkers, single-tenant AI.

My everyday mantra:
“dubito, ergo cogito, ergo sum" - Descartes, Principles of Philosophy (1644)

If you have a problem you can't solve, book a slot to discuss and brainstorm together: https://topmate.io/dexter1618

taplio

The Battle

Adam Sroka

Debayan Das

Followers

Posts

Engagements

Likes

Ready to grow your LinkedIn presence?