All cases
Assistants

Growing health start up gets AI kickstart with rapid AI rollout

Scaling a business becomes much easier with 20 custom assistants.

60%
more accurate research
50
new lead monthly
15 hours
saved weekly

Executive summary

Dendro Health partnered with gecco for a rapid AI enablement programme that converted institutional knowledge into practical, day‑to‑day assistants. Over a short delivery window gecco built and deployed 20 tailored AI assistants, operationalised core documents, and connected those assistants to Dendro’s Notion workspace, email and CRM inside ChatGPT Teams. The programme included hands‑on training in agent mode and deep research techniques, enabling the team to use AI confidently for meeting admin, marketing, sales and strategic research. The engagement delivered immediate productivity gains, faster access to evidence, and a clear capability uplift so the team can extend assistants independently.

About Dendro Health

Dendro Health is a healthcare technology and consultancy organisation that helps clients translate clinical and market evidence into service design, strategy and operational workflows. The organisation combines clinical expertise, specialist research capability and digital tooling to support health systems, providers and life‑science organisations in delivering evidence‑based programmes, bids and service improvements.

Objectives and scope

Objective: Rapidly enable Dendro to use AI assistants that reduce routine work, surface authoritative evidence and integrate seamlessly with existing tools.

Scope: A focused programme covering discovery, deployment of 20 assistants, ingestion and operationalisation of core documents (SOPs, evidence libraries and templates), secure integrations with Notion, email and CRM inside ChatGPT Teams, and practical team training covering agent mode, prompt design and deep research.

Audience: This case study is aimed at healthcare technology teams, research and evidence functions, and consultancies considering rapid AI adoption across knowledge workflows.

Challenges and drivers for change

Before the engagement, Dendro faced several barriers to efficient knowledge use:

  • Fragmented knowledge sources: important documents and data were distributed across Notion, emails and CRM records, slowing discovery and reducing confidence in sources.
  • Manual meeting administration: preparing agendas, capturing notes and following up on actions took significant staff time.
  • Content bottlenecks: marketing and thought‑leadership content took too long to produce consistently.
  • Limited AI experience: the team wanted practical, repeatable ways to use AI rather than abstract demonstrations.
  • Need for secure, governed access: any integrations had to respect data governance and operate inside ChatGPT Teams.

These challenges made a tightly scoped, practical programme the right approach: convert the highest‑value knowledge assets into live assistants and teach staff how to use them safely and effectively.

Strategic goals and vision

Dendro and gecco agreed the programme should:

  • Deliver a working set of assistants that embed into daily workflows within weeks.
  • Ensure assistants answer from Dendro’s authoritative documents so outputs are auditable and trustworthy.
  • Train staff to operate and extend assistants independently (agent mode, prompt design, admin playbooks).
  • Reduce context‑switching by enabling queries across Notion, email and CRM from a single chat environment.

The longer‑term vision was to move from experimentation to routine, sustained use — making evidence instantly accessible to the people who need it.

Solution and approach

gecco delivered a tight, four‑phase programme designed for speed and durability:

  1. Rapid discovery (workshops): Short, focused sessions mapped Dendro’s workflows, prioritised the highest‑value assistant use cases (meeting admin, marketing, sales, strategy), and identified the core documents and CRM fields to index.
  2. Onboarding & deployment: gecco built and deployed 20 assistants including meeting‑admin agents (agenda creation, minutes, action capture), marketing assistants (drafts, outlines, event materials), CRM lookup and outreach assistants, and strategic research assistants for evidence synthesis and market scans. Core documents and templates were ingested or indexed so assistants answer from Dendro’s sources.
  3. Integrations: Assistants were connected to Notion, email and the CRM inside ChatGPT Teams, enabling contextual lookups and secure access without leaving the chat interface. Access controls and admin playbooks were put in place to maintain governance.
  4. Training & handover: Hands‑on workshops taught agent workflows, deep research methods (source triangulation, evidence extraction), prompt design and iteration. gecco supplied admin playbooks and documentation so Dendro could maintain and extend assistants.

Throughout implementation gecco prioritised quick wins: the highest‑impact assistants were delivered first so users saw immediate value and helped refine later builds.

Results and outcomes

The programme produced immediate, practical outcomes:

  • 20 live assistants deployed and in routine use inside ChatGPT Teams.
  • Core documents operationalised within assistants so answers are grounded in Dendro’s evidence and policies.
  • Integrated workflow: Notion, email and CRM are now queryable from the chat environment, cutting time spent switching tools.
  • Capability uplift: staff learned agent mode and deep research techniques and now extend assistants using gecco’s playbooks.

Functional improvements by category:

  • Meeting admin: agendas, minutes and action lists are auto‑generated, reducing administrative time and improving follow‑through on actions.
  • Marketing: draft blogs, newsletters and social outlines are produced rapidly, increasing campaign velocity and consistency.
  • Sales: CRM‑linked assistants provide instant lead context and produce personalised outreach, improving responsiveness and pipeline hygiene.
  • Strategy & research: rapid evidence synthesis and market scanning shorten the time from request to insight, supporting faster bid preparation and decision‑making.

Qualitative feedback from the team highlighted faster onboarding for new starters, more consistent evidence outputs, and greater confidence in daily use of AI tools.

Lessons learned

  • Prioritise assistants that map to everyday tasks. Rapid adoption follows clear, repeatable value (meetings, CRM lookups, content drafts).
  • Train with real work. Hands‑on training using actual tasks accelerates learning and embeds habits.
  • Governance by design. Building integrations inside ChatGPT Teams with clear admin controls made the team comfortable using assistants with live data.

Get in touch

Want rapid, secure AI enablement for your team? gecco specialises in short, high‑impact programmes that turn knowledge into useful assistants — from discovery to deployment and training.

Book a demo
Training & Consultancy

Five-day training unlocks AI consistency

Deduxer, a digital agency partnered with gecco for a five day training and consultancy programme.

Agents

Hands off meeting preparation

Revcrank successfully deployed an end-to-end AI agent that manages their entire meeting workflow.