Proposal — Submitted May 12, 2026
Practical. Hands On. Secure.

Responsible AI for
Echidna Giving

Helping your team use AI well — with a shared policy, the right tools, and the confidence to put them to work.

Prepared by Shabnam Aggarwal Project kickoff June 4, 2026

Shabnam brought an exceptional combination of technical expertise, strategic insight, and adaptability to our work. She led a series of workshops that helped our team think more intentionally about how AI could support our grantmaking, while also developing and training a due diligence agent tailored to our processes and priorities. What was particularly remarkable was how quickly she absorbed our team's highly specific ways of working and incorporated feedback in real time — it genuinely felt as though we had gained a new teammate with deep AI expertise.

Kelsey Credé  •  Executive Director, Robertson Foundation

Most of your team is already
using AI. The foundation hasn't been built yet.

On our May 5th call, your team was specific about what matters. This proposal is organized around those priorities.

0
%
of your 15-person team uses AI tools at least twice a week.
No shared policy. No agreed standards. Everyone flying blind in cloudy skies.

Privacy is the top concern

You hold sensitive grantee, funder, and partner data. Staff need clear guidance on what can and cannot go into any AI tool, and under what conditions.

Accuracy and bias

AI-generated content needs quality guardrails, especially in grantmaking contexts where the quality of analysis and recommendations has real consequences.

Reputational risk

How AI use appears to partners, grantees, and funders matters. A documented, deliberate policy signals intentionality — not just adoption for its own sake.

Staff overreliance

The worry isn't that staff will use AI. It's that heavy reliance could erode the quality of analysis and portfolio support — especially without shared standards for when to trust AI output and when to push back.

Echidna's situation is not unusual. Across organizations, the gap between individual AI use and organizational readiness is widening — and the cost of that gap is measurable.

67%

of AI outcomes at work are driven by organizational factors — leadership alignment, manager support, talent practices — not individual skill. Even the most AI-fluent employee captures only half the value without organizational backing.

47%

of nonprofits have no AI governance policy at all — even as 92% report using AI in some form. Organizations without governance are almost never among those reporting meaningful improvements in capability.

40%

of potential AI productivity gains are left unrealized when organizations lack a structured training strategy. Only 12% of employees say they receive AI training they consider actually sufficient.

Organizations that build structured governance and training early are four times more likely to report major improvements in organizational capability than those in ad hoc adoption — and the gap is widening. Only 26% of AI users say their leadership is consistently aligned on strategy. The bottleneck is not the tools. Grant Thornton 2026 AI Impact Survey ↗

Three phases. June through September.

The engagement is built around three phases, each closing with something your team can act on before the next begins. Working group sessions are built into every phase.

1
Jun 4 — Jun 20 • 2 weeks

Understand the Real Picture

Before any recommendation gets made, we need to know what's actually happening. Phase 1 opens with the Organizational AI Diagnostic — a structured individual assessment run with each member of your working group. It surfaces where staff are in their AI fluency, what assumptions they're carrying, and where the team's positions diverge in ways that matter for policy. The diagnostic has been run with peer foundations in your sector and consistently surfaces things that group sessions miss.

  • Kickoff session with the AI working group
  • Organizational AI Diagnostic administered to working group members individually
  • Analysis of existing staff survey data alongside diagnostic findings
  • Targeted review of EGGs (SmartSimple) and Google Workspace
  • Conversations with key staff across programs, operations, and leadership
  • Deliverable: Prioritized use case assessment with risk map, grounded in diagnostic findings
✦ The Organizational AI Diagnostic has been piloted and refined through engagements with foundations in the education and international development sectors. Echidna's working group receives a version tailored to your grantmaking context and data environment.
2
Jun 20 — Aug 1 • 5 weeks

Build the Foundation

Translate the use case assessment into three durable assets: a tool recommendation, an adoption strategy, and a living AI policy your team can enforce and update as the landscape changes.

  • Evaluation of AI tools suited to Echidna's data sensitivity and team size
  • Recommendation on enterprise solution, commercial LLM, or hybrid approach
  • AI policy and data governance documentation, written as a living document
  • Risk management framework including vendor agreement guidance and data handling
  • Deliverables: Tools recommendation, adoption strategy, organizational AI policy, risk framework
✦ Shabnam will be in the Bay Area during Phase 2 and is available for in-person working sessions. Choosing in-person delivery adds approximately $4,000 to the Phase 2 budget. Please flag your preference in your selection letter.
3
Aug 1 — Sep 30 • 8 weeks

Bring the Team Along

Your staff have a wide range of comfort levels with AI tools. Training is designed starting from where people actually are, across roles and technical backgrounds. This phase also includes implementation support as priority use cases move from paper to practice.

  • Training program design tailored to role-based needs across your team
  • Two to three facilitated training sessions, in-person or virtual
  • Implementation support for 3 to 5 priority use cases during rollout
  • AI tool adoption decision framework — a reusable go/no-go evaluation guide for the working group as new tools emerge
  • Grantee-adaptable version of key policy sections, available for optional sharing
  • Deliverable: Training materials, facilitated sessions, decision framework, implementation wrap-up
Key dates
Kickoff session with the AI working group. Scope confirmed, discovery begins.
Jun 4
Project kickoff
Use case assessment delivered. Working group alignment session before Phase 2 begins.
Jun 20
Phase 1 deliverable
Policy, tools recommendation, adoption strategy, and risk framework delivered.
Aug 1
Phase 2 deliverable
Training complete. All materials handed over. Final wrap-up session.
Sep 30
Project close

Six deliverables, mapped directly to your RFP.

Where grantee-adaptable versions of materials make sense, those are included — available for Echidna to share at its discretion, at no additional cost.

Deliverable 1

AI Use Case Assessment

Grounded in the Organizational AI Diagnostic — a structured individual assessment run with each working group member before any group sessions begin. Surfaces real fluency levels, hidden assumptions, and where staff positions diverge. Produces a prioritized use case map with risk analysis, aligned with senior leadership.

Deliverable 2

AI Tools Overview & Recommendation

Evaluation of tools relevant to your priority use cases. Benefits, risks, and data security profile for each. A clear recommendation with rationale.

Deliverable 3

Adoption Strategy & Decision Framework

Guidance on commercial, enterprise, or hybrid approaches. Includes 3 to 5 high-value, low-risk use cases for immediate adoption, plus a go/no-go evaluation framework your working group can use independently when new tools emerge.

Deliverable 4

Organizational AI Policy

A living policy document covering acceptable use, data governance, and role-based guidance. Defines clearly which data can and cannot be used with which tools, and under what conditions.

Deliverable 5

Risk Management Framework

Safeguards for data ownership, storage, and vendor accountability. Guidance on vendor agreements, insurance considerations, and monitoring for a team without a dedicated tech function.

Deliverable 6

Training Program & Facilitation

Role-appropriate training for your growing team. Facilitated sessions delivered directly. Materials built for reuse as the team grows and the policy evolves.

Two sides of the table.

Most consultants in this space come from one direction: they have either worked inside philanthropies or alongside grantees. Shabnam has done both, at the size and type of organization that matches this engagement. She led Rising Academies' AI integration work as well as Robertson Education team's. Rising is a current Echidna grantee. She knows what responsible AI adoption looks like from where you sit and from where they sit. That double vantage point is what this kind of engagement requires.

Shabnam Aggarwal
Lead Consultant

AI product builder, government AI advisor, and consultant to foundations on responsible AI adoption. Built Rori, an AI-powered learning tool considered a model in the sector across Sub-Saharan Africa. Currently advising Gates Foundation and Mulago on AI strategy on exactly this type of work.

  • AI policy, diagnostic, and training engagements with Mulago, Robertson, and Gates Foundations
  • Current AI strategy advisory work with Gates Foundation
  • Led AI integration for Rising Academies, a current Echidna grantee
  • Built AI products for government and development sector clients
  • Familiar with California AI regulatory context and federal governance frameworks
  • Author and speaker on responsible AI adoption in resource-constrained organizations
HCD & Policy Researcher
Research & Policy Support

A human-centered design and policy researcher with direct experience in nonprofit AI governance, data security documentation, and training design for mission-driven organizations. Brings structured research methodology to discovery and helps translate complex guidance into materials staff can actually use.

  • Experience in philanthropy and international development sector AI projects
  • HCD-based discovery: surveys, interviews, synthesis across distributed teams
  • Policy drafting, data governance documentation, and training content development
  • Embedded across all three phases without adding coordination overhead to your team
Is this the right fit?

This engagement is designed for a foundation at Echidna's size and stage — a small, mission-driven team where AI is already in use but governance hasn't kept up, and where staff capacity is the actual goal, not a deliverables binder on a shelf.

It's a strong fit for organizations that want someone who understands how foundations make decisions and how grantees experience their funders' AI choices. Someone who can work directly with staff at every level, brings fluency from actually building AI products, and isn't selling a stack.

If your primary need is large-scale IT implementation, a multi-person firm with dedicated legal and technical bench depth, or full-time on-site presence for the duration, this engagement isn't the right structure — and Shabnam will tell you that directly rather than over-scope to fit your budget.

The budget, by phase.

Four months, June through September.

Phase What this covers Deliverables Investment
Phase 1
Jun 4 – Jun 20
Kickoff session, staff survey analysis, systems review (EGGs + Google Workspace), stakeholder conversations across programs and ops, use case mapping Prioritized use case assessment with risk analysis
$13,500
Phase 2
Jun 20 – Aug 1
AI tools evaluation, adoption strategy, policy drafting and iteration, data governance guidance, risk management framework, working group sessions Tools recommendation, adoption strategy, organizational AI policy, risk management framework
$26,700
+$4,000 for in-person Bay Area workshops (optional)
Phase 3
Aug 1 – Sep 30
Training design, facilitated staff sessions, implementation support across priority use cases, AI tool adoption decision framework, grantee-adaptable materials, final handover Training program and materials, decision framework, facilitated sessions, implementation wrap-up
$9,800
Total Jun 4 – Sep 30  •  4 months 6 deliverables
$50,000

AI's changelog moves faster than
any policy document.

Echidna currently has no in-house technical resource tracking AI developments. The tools, regulations, and best practices your team relies on in September will already be shifting. This optional retainer keeps the guidance current and gives staff somewhere to bring questions as real-world use evolves.

Monthly Advisory

Technical Check-in

$2,000 / month

A focused monthly session to review AI developments relevant to Echidna's work and address any questions your team has raised since the last conversation.

  • Monthly session with Shabnam
  • Policy and tool updates as the landscape shifts
  • Staff questions handled between sessions via email
  • Written summary of key developments after each session

Minimum 3-month commitment after project close, rolling monthly thereafter. Can also be engaged mid-project if the team wants ongoing support before September.

Questions? Book a call.

Happy to connect whenever you have questions. Use the link below to find a time.

Schedule a call with Shabnam