Fractional Chief AI Officer, &
Autonomous Systems Operator.
Schedule a CAIO Strategy Call
Direct architecture consultation with Harshit to evaluate agent feasibility and calculate ROI.
Who This Is For
Explorer, Idea Stage, SME, or Enterprise — each gets a different engagement model, not the same pitch.
Explorers
Join CohortLearning to build AI-native systems hands-on, before committing to a company or a role.
Idea Stage
PartnerYou have the idea and the drive, not yet the product. I partner in for equity, not invoices.
SME
ContractFunded and shipping. Fixed-scope audits, production sprints, and embedded fractional CAIO retainers.
Enterprise
Hire MeNot currently taking full-time roles — open to the right conversation if one comes along.
Choose Your Fractional CAIO Partnership Mode
From high-velocity 1-week stack & threat audits to fixed-scope 2–4 week build sprints and monthly embedded executive leadership.
Select Your Fractional CAIO Partnership Mode
From high-velocity 1-week stack audits to full-scale multi-agent build sprints and monthly executive retainers.
Production Agent Sprint
End-to-end hands-on engineering sprint embedding directly into your codebase. Architect, test, and deploy production-grade autonomous agent runtimes or sovereign ops pipelines.
What You Receive
Zero spam. Direct line to Harshit (IST / Bangalore).
Have the Idea, Not Yet the Product?
A small number of idea-stage founders get a technical co-builder for equity, not an invoice.
You have the conviction and the idea, not yet the product or the funding to hire. I partner in as a technical co-builder for a small number of idea-stage founders at a time — equity in exchange for the same production-grade engineering that funded SME clients pay for.
Schedule a Partnership Call
Tell me about your idea. If it's a fit, we discuss an equity-for-work partnership, not an invoice.
Where I’ve Worked, and What Changed
Product work at Lyzr AI and PhonePe, zero-to-one builds with founders, and teaching — each with the numbers it moved.
Lyzr AI
/ Member of Technical Staff · Product ManagerLyzr ADK Rollout, Multi-Agent Templates, Partners Programme & Lyzr University
Product work on the core platform of a model-agnostic, enterprise-grade agent platform: rolling out the Lyzr ADK to developers, building the multi-agent template generator, running the Partners Programme, and setting up Lyzr University.
Getting developers, partners and learners to adopt an enterprise agent platform, alongside the existing Lyzr Studio.
Led the Lyzr ADK rollout over Lyzr Studio, built the template generator for multi-agent systems, ran the Partners Programme end to end with SIs, GSIs and consultancies, and set up Lyzr University and its Thinkific academy.
PhonePe
/ Associate Product Manager I → IIPayment Gateway: Merchant Dashboard, Refunds & Notifications
Product manager on the payment gateway merchants use to collect payments: the payer interface and post-live merchant experience as APM I, then Payments-as-a-Service billing visibility and merchant settings as APM II. Promoted within a year.
Merchants raised avoidable queries about transactions, bulk refunds failed on repeat errors, and pending merchant actions went unnoticed.
Improved transaction filters and attributes, optimised the bulk refund flow, built notification workflows for pending merchant actions, drove a migration to current platform standards, and wrote SOPs and documentation for the support teams.
NxtJob.AI
/ Product ManagerAI Product Strategy for a Job-Search Platform
Aligned AI product and feature strategy with business and revenue goals across the resume builder, interview prep and job-board products, and set up product practices that increased feature delivery by 25%.
An early-stage AI job-search platform needed its AI features tied to business goals, and a steadier way to ship them.
Mapped AI features to business and revenue goals, set up product practices from discovery to delivery, and helped shape the team around the roadmap.
Project Dark Horse
/ Product Manager & Startup-School InstructorIdea to Concept MVP with 3 Startups, Plus a PM Curriculum
Worked with 3 idea-stage startups at a micro venture studio, taking each from idea to concept MVP on a 30-day turnaround, and wrote and ran the product management curriculum for its startup school.
Idea-stage founders needed to get from an idea to something testable quickly, and the startup school needed a product management curriculum.
Worked alongside each founding team from idea to concept MVP, and wrote and taught the PM curriculum for the cohort.
Agile Academy
/ ICAgile-Accredited PM InstructorTeaching AI-Native Product Management to 100+ PMs and Operators
Taught discovery to delivery with AI-native product skills to 100+ product managers and engineering operators in cohort-based courses, and built a sandbox externship where learners worked through 0→1 and 1→n cycles and shipped working MVPs with AI features.
Learners needed practice, not just frameworks: a safe place to run real product cycles with AI tools.
Taught the ICAgile product management curriculum with AI-native skills, and built a sandbox externship that simulates 0→1 and 1→n product cycles.
BricknBolt
/ Product ManagerContractor & Homeowner Lifecycle, Approvals and a New Works Channel
Owned the central platform for construction project financials, resources and timeline management; simplified approvals across the value chain for a 6% TAT improvement; and worked on a new ‘works’ channel that brought in ₹30 L MRR (0.3% of total).
Approvals across homeowners, contractors and project teams slowed construction milestones.
Simplified the approval experience across the value chain, and shaped the contractor and homeowner lifecycle across the app, WhatsApp and other channels.
What Every Engagement Ships
No vague slide decks or bloated billable hours. Every engagement delivers production-ready code in your repo, hardened schemas, and enabled teams.
What Your Company Receives
No vague advisory decks. Every Fractional CAIO engagement ships 100% owned source code in your repository, hardened schemas, and enabled internal teams.
Agent Topology & Domain Schemas
Complete architectural specification of multi-agent state machines, relational database schemas, tool boundaries, and security invariants.
Model-Agnostic Agent Runtimes
Working agent code in your GitHub repo, built on the framework that fits your stack.
Human-in-the-Loop Control Planes
Interactive operator interfaces and 1-click Slack/web approval gates that empower your team to oversee autonomous agent actions.
Private Runbooks & Team Coaching
Hands-on enablement sessions and comprehensive documentation to ensure your internal developers and operators maintain the stack independently.
The 4-Stage Architecture Protocol
A battle-tested engineering methodology for mapping, architecting, deploying, and governing sovereign agent infrastructure.
How We Architect & Ship Sovereign Systems
A disciplined, low-risk engineering methodology for deconstructing legacy drag, building model-agnostic agent runtimes, and enabling internal developers.
Agent Stack & Threat Diagnostic
Deconstruct existing prompt chains, model dependencies, SaaS sprawl, and tool sandboxing vulnerabilities across your engineering stack.
Deterministic Architecture & Schemas
Design robust multi-agent state machines, strict JSON schema guards, automated eval test harnesses, and isolated tool execution bounds.
Repo-Native Build & Verification
Embed directly into your GitHub repository to write, test, and ship production-ready agent loops with automated fallback handlers.
Team Enablement & CAIO Governance
Conduct 1-on-1 team architecture walkthroughs, deliver operational SOP runbooks, and provide weekly PR review stewardship.
How a Fractional CAIO Compares with a Full-Time Hire or an Agency
A part-time engagement, side by side with hiring full-time or bringing in an agency.
Why a Fractional CAIO Beats Full-Time Hires & Generic Agencies
How a part-time engagement compares on cost, speed and ownership.
Fraction of executive burn. Zero equity dilution, zero payroll tax overhead.
A full salary and equity, which is a lot to carry at an early stage.
Unpredictable hourly billing, change-order creep, and opaque junior developer markups.
Immediate architecture review, deterministic sprint roadmaps, and rapid PR merges.
Lengthy executive search, notice periods, and 90-day onboarding ramps.
Heavy account management overhead and slow feedback loops with off-shore teams.
Pure TypeScript/Python state machines, open schemas, zero vendor lock-in or SaaS taxes.
High key-person dependency if the executive departs after building custom infra.
Proprietary scaffolding, obfuscated logic, and ongoing maintenance retainers.
Model-agnostic agent runtimes, isolated tool sandboxes, and idempotent state machines.
Often theoretical AI strategy with limited hands-on modern agent compilation speed.
Brittle prompt chains prone to hallucination, tool leaks, and silent production crashes.
Recorded 1-on-1 walkthroughs, automated test harnesses, and SOPs for internal team autonomy.
Builds internal org, but creates single-point-of-failure governance.
Engineered for ongoing dependency, keeping your internal operators in the dark.
Executive FAQ
Everything you need to know about timelines, intellectual property, tech stack compatibility, time commitment, and pricing models.
Frequently Asked Questions
Everything executive teams and founders need to know about timelines, IP ownership, stack compatibility, and time commitment.