Current application narrative. Use this page for 2026 applications. Old Stumble answers are preserved in the archive and should not be reused without rewriting.

Source-of-truth profile

Founder: Sarthak Agrawal
Stage: Solo technical founder; employed; multiple AI/devtools prototypes; pre-revenue and pre-design-partner unless updated below
Experience: Approximately four years building and operating production software across Go, Kafka, TypeScript, distributed systems, fintech workflows, and AI-enabled products
Current exploration: Verification, observability, and operational control for software increasingly built and maintained by AI agents
What is missing: Concentrated problem selection, customer evidence, distribution, and sustained external accountability—not implementation ability
Program preference: Apply when a program offers meaningful funding, brand, mentorship, network, structure, community, customers, or relevant credits. Prefer to leave the job only after a signed offer provides meaningful personal support or enough upfront company capital to create responsible runway.

One-line descriptions

Founder

Product-minded software engineer who has repeatedly shipped production systems and independent AI products, now concentrating on the verification and operational-control layer for agent-built software.

Company thesis

We give teams evidence that AI-generated software changes actually satisfy intent and keep real applications healthy.

Expanded product thesis

Coding agents are making implementation cheap, but generation is improving faster than verification. The company would observe what users actually experience, identify consequential failures, give coding agents the evidence needed to fix them, and independently verify recovery.

Founder bio

Approximately 50 words

I am a product-minded software engineer with approximately four years of experience across backend systems, event-driven infrastructure, fintech workflows, and AI products. At Vault Wealth, I have contributed to production systems as usage scaled substantially. Independently, I repeatedly turn ambiguous ideas into working software across AI verification, agent infrastructure, and local models.

Approximately 120 words

I am a product-minded software engineer with approximately four years of experience building and operating production software across Go, Kafka, TypeScript, distributed systems, fintech workflows, and AI-enabled products. At Vault Wealth, my strongest work has involved owning more than implementation: understanding a user problem, designing the system, shipping it, and measuring whether it changed behaviour. Outside work, I repeatedly build ambitious products without needing external direction, including systems for AI code verification, local model post-training, agent infrastructure, and application operations. My advantage is the combination of product reasoning and systems execution. My current failure mode is not an inability to build; it is spreading that capability across too many technically interesting directions before proving external demand.

Core application answers

▸ Why do you want to start a company?
▸ What are you exploring?
▸ What might the initial product be?
▸ What is your edge?
▸ Why now?
▸ What is your biggest weakness as a founder?
▸ What do you want from a program?

Commitment answer

I am applying while employed, and applying does not imply that I have pre-committed to resign. I can participate immediately in programs compatible with a full-time job. For an intensive or full-time residency, I would make the employment decision after seeing the actual offer, people, location, program structure, financial terms, and visa logistics. My preference is to leave only when meaningful personal support or enough upfront company capital creates responsible founder runway, but I remain open to exceptional brand, community, or mentorship opportunities even when their acceptance decision requires a separate trade-off.

Why apply before deciding whether to accept?

An application creates information and optionality. A program can be worth applying to for brand, peers, mentors, network, community, credits, or access even when its full-time format may not ultimately fit. I would make the final commitment decision only after receiving concrete terms and understanding the cohort, attendance, economics, and opportunity cost.

Cofounder answer

I can build independently and am not waiting for a cofounder, but I am open to forming a complementary founding team when the partnership materially compounds judgment, distribution, domain access, and ambition. I would not add a cofounder for accelerator optics.

Program-specific answers

▸ Why SPC?
▸ Why Antler Korea?
▸ Why Conviction Embed?
▸ Why Afore FIR?
▸ Why Entrepreneur First?
▸ Why GSF Founders Academy?
▸ Why Founder Institute Hyderabad?
▸ Why Z Fellows?
▸ Why Antler India Before Day Zero?
▸ Why Athena Ventures?
▸ Why One Founder Fellowship?

Evidence and metrics — update before every submission

Claims requiring verification before use

Exact personal ownership of any 15K-to-200K DAU scaling claim
Exact basis and attribution of any 40% engagement improvement
Exact basis and attribution of any 90% support-load reduction
Current user, revenue, retention, and repository-adoption figures
Full-time availability and resignation timing
Relocation and visa feasibility for each location

Reusable submission packet