Add senior AI judgment without adding another executive seat.
When AI decisions start affecting strategy, vendors, architecture, investment, and operations at the same time, isolated advice is no longer enough. I work alongside founders, executives, and technical leaders on the decisions that are expensive to reverse: what to pursue, what to buy, what to build, how to architect it, what risk to accept, and what should wait.
Recommended 90-day working horizon. No minimum term.
If you are carrying AI decisions that are expensive to reverse, this is for you.
You do not need a specific title. You need to be the person who cannot afford to get the AI decision wrong and cannot afford to delay it either.
The founder deciding what to build first
You have limited capital, limited time, and multiple AI opportunities competing for both. You need someone to help you sequence the work and commit without overcommitting.
The executive approving the spend
You are being asked to fund AI initiatives, sign vendor contracts, and approve architectures. You need an independent view before the money moves, not after.
The CTO deciding build, buy, or wait
You are responsible for the architecture, the data, the security boundaries, and the operational reality. You need someone who speaks both engineering and executive and can pressure-test the plan.
The operator crossing from experiment to production
Your organization is moving from AI experiments to AI that has to work. The stakes change. The decisions change. You need someone who has been through that transition.
The leader in a regulated industry
Regulations, security requirements, and governance constraints are affecting your architecture and your rollout. You need someone who can connect compliance to execution without paralyzing either.
The team leading multiple AI initiatives at once
Different business units are pursuing disconnected AI work. You need someone to see the whole portfolio, identify the conflicts, and help leadership prioritize across all of it.
If you are the person who has to answer "what do we do about AI?" and the answer actually matters, this is for you.
Concrete situations where this relationship is worth it
Several AI initiatives compete for attention and leadership needs prioritization.
A vendor or platform choice creates a long-term dependency before the contract is signed.
Architecture and business priorities conflict and someone needs to translate between them.
The roadmap sequence is unclear and leadership needs pressure-testing before committing.
An investment decision is expensive to reverse and needs an independent view first.
Governance and execution need to be connected, not documented and forgotten.
The organization has crossed from experimentation into operating AI.
Regulations or security requirements affect architecture or rollout decisions.
Multiple business units are pursuing disconnected AI work.
Leadership needs an independent point of view before approving spend.
None of these require a crisis. They require decisions that are expensive to reverse and benefit from independent senior judgment before commitment.
The decisions this relationship covers
Strategy
Where AI can materially change the business and what should not be pursued.
Opportunity Portfolio
Identify, compare, and sequence opportunities. Not all opportunities deserve investment.
Investment Decisions
What deserves capital, people, and executive attention. What does not.
Vendor Evaluation
Vendor capability, claims, architecture, data, economics, dependency, and exit path.
Build / Buy / Configure / Connect / Wait
The full decision set, not a binary build-versus-buy. What to buy, build, configure, connect, or wait on.
Architecture
Systems, data flows, integrations, agents, permissions, security boundaries, and operational design.
Roadmaps
Sequencing, dependencies, evidence gates, and stop conditions. What each phase must prove before the next begins.
Operating Model
Who decides, owns, approves, monitors, and can stop AI systems. Centralized, federated, or hybrid.
Governance and Regulatory Exposure
Where legal, regulatory, security, privacy, or governance constraints materially affect the decision. Not legal advice.
Implementation Sequencing
Translate decisions into executable next steps without automatically becoming the implementation team.
The product is not the meetings. The product is judgment, continuity, and an operating record.
Two monthly working sessions are the visible cadence. The actual value is what happens between and across them: persistent context, senior independent judgment, decision continuity, pressure-testing, evidence, architecture and business translation, decision artifacts, and an operating record of what was decided and why.
Your strategy, systems, vendors, decisions, and constraints stay on file. I do not restart from zero each session.
Independent review from someone who has shipped production AI systems and can tell you whether the plan survives contact with reality.
Context carries across decisions. What we decided last month affects what we recommend this month.
I challenge the first answer. The stated problem is often not the problem worth solving.
Recommendations are backed by evidence, not assertion. What supports the decision. What weakens it.
I speak both languages and can translate between the engineering team and the executive team.
Selected written artifacts tied to active priorities. Not unlimited document production.
A record of what was decided, by whom, why, and what happened next. Your executives can defend it without me in the room.
The method behind every decision
The same six disciplines, adapted to an ongoing advisory relationship. Not over-branded. Connected to actual executive decisions.
Research
I understand your current systems, the decisions in flight, and your constraints. No generic playbooks.
Reframe
I find the real question behind the request. The stated problem is often not the problem worth solving.
Prove & Decide
We work the options, the evidence, and the trade-offs. You leave with a recommendation an executive can sign.
Architect
I design the system, data, human, and control model behind the decision. Not just what to do, but how it operates.
Mobilize
We clarify ownership, dependencies, risks, and cadence so the decision has named owners and visible next steps.
Improve
I use production evidence to scale, change, or stop. The loop closes when the decision meets reality.
What the workspace actually holds
The workspace is not a shared drive. It is a structured operating record that moves with the work. Every decision passes through the same loop you saw above: Context to Opportunities to Decisions to Evidence to Commitments to Outcomes.
What exists inside the workspace
Artifacts are tied to active priorities and scope. This is not unlimited document production.
What the work produces
Not every artifact every month. The artifacts that match the priorities we are working through.
Decision Brief
A focused written analysis of one decision: options, evidence, trade-offs, and recommendation.
Vendor Comparison
Side-by-side evaluation of vendors against your requirements, not a generic checklist.
Architecture Review
Review of a proposed system with security, data, and operational considerations.
Roadmap Review
Pressure-test of sequence, gates, and evidence thresholds.
Decision Record
A written record your executives can defend without me in the room.
Risk / Assumption Register
What could go wrong, what we are assuming, and what would change the recommendation.
Operating Recommendation
Who owns what, how decisions get made, and how the system gets stopped if it fails.
Opportunity Prioritization
Ranked comparison of opportunities by value, fit, feasibility, and risk.
Implementation Sequencing Recommendation
What to build first, what to test, what to deploy, and what to measure.
Why have me working through these decisions with you?
I do not approach AI only as a strategist, a developer, or a governance specialist. I work across the decision.
Does the opportunity create enough value to deserve attention?
Can it work with the systems, data, and operating environment you actually have?
Should you buy, configure, connect, or build?
Who owns it? What happens when it fails? Where does a human remain involved?
What data, security, regulatory, compliance, or accountability exposure are we accepting?
How will we know whether the decision worked?
Verified proof behind the work:
- Enterprise application portfolio leadership
- 83+ projects delivered across enterprise, client, and founder-led work
- Production AI systems, including Kestrel Voice and HAIEC
- Systems architecture and governance and compliance work
Fractional AI Advisor
Fractional AI Advisor
$1,250/monthStrategy, opportunity, vendor decisions, architecture, roadmaps, governance, and operating model. Month-to-month. No minimum term.
What the relationship includes:
Annual: $12,500/year (twelve months for the equivalent of ten monthly payments). Monthly is the primary choice. Annual is available if you prefer it.
Recommended working horizon: 90 days
The engagement is month-to-month. I recommend starting with a 90-day working horizon because the value comes from carrying context across several decisions, not from one isolated meeting.
At approximately 90 days, a structured review:
- Decisions made
- Opportunities advanced or rejected
- Risks surfaced
- Assumptions changed
- Outcomes and evidence
- Priorities for next period
- Whether Fractional is still the correct relationship
You can continue, move to AI Advisor Desk, expand scope, or stop.
How the relationship starts
After payment, you complete a deeper Context Intake than the $99 plan. Then you schedule the Activation Call and your first Working Session directly from the workspace.
Context Intake captures:
Schedule your Activation Call
20 minutes complimentary
- Validate priorities
- Confirm immediate decisions
- Establish relationship cadence
- Clarify stakeholders
- Confirm first working session focus
Self-scheduled from the workspace. Timezone aware. Does not count against your two monthly sessions.
Schedule your first Working Session
60 minutes
The first real working session on your priorities. Can be scheduled after the Activation Call or simultaneously if operationally sensible.
Self-scheduled from the workspace. Timezone aware. Cancellation and rescheduling available.
This is one relationship in a larger model
Need a lighter relationship?
If you primarily want someone watching what changes, maintaining context, and helping with focused AI decisions as they arise, AI Advisor Desk starts at $99/month.
Need implementation, not advisory?
Fractional AI Advisor is advisory. If a decision has been made and you need substantial architecture, implementation, integration, or build work, that becomes a separately scoped engagement. I flag this before implementation work begins. The Fractional retainer is not silently consumed by build work.
Member tools included with your subscription
Human judgment comes first. These supporting tools extend what the advisory relationship can do. Product limits apply to each included tool.
HAIEC SCAN
Kestrel AI Number Basic
Member Tool Library
Included product access is a supporting capability, not the reason to subscribe. HAIEC and Kestrel maintain their own product limits, acceptable-use terms, and upgrade paths. Advisory subscription does not create additional uptime or SLA guarantees for those products.
How the engagement works in practice
Advisor affiliation
An active Fractional client may, with approval, identify Subodh KC as Fractional AI Advisor or External AI Advisor in appropriate materials.
This does not create employment, officer status, agency, fiduciary authority, or authority to bind either party. Public quotes, press releases, logos, or alternative titles require approval. Full terms are in the Fractional AI Advisor Service Terms.
Advisory service expectations
Acknowledgment
Priority asynchronous requests are normally acknowledged within one business day.
Substantive response
Substantive response normally within two business days when reasonably within Fractional advisory scope.
Not 24/7 support
Fractional AI Advisor is not 24/7 support, managed incident response, or emergency technical support.
System of record
The workspace remains the system of record. Working sessions are scheduled calls.
Session policy
Common questions about fractional AI advisory
Start Fractional AI Advisor
Senior AI judgment working through your strategy, vendor, architecture, roadmap, and governance decisions. Month-to-month. Recommended 90-day working horizon.
Who is this Advisor Desk for?
Enter the organization or business you want this Advisor Desk to support.