Skip to main content

AI product builder · Engineering leader

AI products people trust.

I turn ambiguous opportunities into working AI products—combining product discovery, hands-on engineering, and rigorous evaluation.

Sunny Patneedi, smiling, in a dark jacket in front of colorful blurred lights
Sunny PatneediFounder of SayMake
20 years
building software, data, and ML systems
1B+
users reached across Apple and Microsoft products
  1. Earlier

    Microsoft · Salesforce

    Helped build products and intelligent systems.

  2. 7 years

    Apple

    Led three engineering teams building large-scale data and ML systems.

    WWDC tentpole features launched Transit cards & fares · EV routing · CarPlay · 3D Maps

  3. Now

    SayMake

    Founder. An AI-supported maker studio for young people, in private preview.

    Building with Agentic harness loops · pgvector retrieval · Runtime security checks

In this collection

SayMake

Kids who build real things, and can explain how.

SayMake gives young builders a project to make, test, explain and improve, with AI support along the way.

For
Families with young builders
Stage
Status: Private preview
Web App
Details
Built with
React · Lexend + Inter design system · Express 5 · PostgreSQL · Drizzle. Full stack and checks
Launch film
Where it stands
Checked 14 October 2026. The public site loaded. It says access is by application, with a waitlist for the coached cohort. The free family shelf reached its sign-in step. I did not complete sign-in.

Rally

New hires who can make the calls your experts make.

Rally turns your team’s reviewed knowledge into a plan for each role, practice on realistic scenarios and AI coaching, with feedback from people kept separate.

For
New hires and the teams who onboard them
Stage
Status: Experiment
Web App
URL
Details
Built with
React · Vite · Tailwind · Express 5 · PostgreSQL. Full stack and checks
Launch film
Where it stands
Not checked yet. A release candidate shown with synthetic company demos. No public page has been checked, it is not a verified production launch, and there is no evidence yet that it shortens ramp-up time.

Cue

Hand work to any AI model and know what it actually did.

Cue, short for Common User Environment, gives every model you use the same permission boundaries and the same evidence of what it tried and what it finished: Claude Code and Codex today, an optional Apple model on the Mac, with Grok, Gemini and fine-tuned open-source models planned.

For
Builders who hand real work to more than one AI model
Stage
Status: Experiment
Web App
URL
Details
Built with
SwiftUI · macOS 15 · iOS 18 · Local Codex · Connected Claude Code. Full stack, checks and agent system
Launch film
Where it stands
Not checked yet. Experimental. Local testing only; reliability work is ongoing. No public page has been checked.

SPEAR

Engineers who can test their LLM app against attack.

SPEAR is open-source middleware that enforces prompt security at runtime, with a red-team harness for trying attacks against it.

For
Engineers building LLM applications
Stage
Status: Experiment
Open Source
Details
Where it stands
Checked 9 October 2026. The public repository, source tree, and README loaded. This supports View source only. It is not a released package, hosted product, or supported workflow.
Earlier capture of the SPEAR repository and evaluation documentation

CareStellar

Families who bring the whole health story to every appointment.

CareStellar is a free prompt and skill for the AI assistant you already use: it organizes the records you have into one summary, with sources, dates and open questions kept visible. Not medical advice.

For
Families caring for someone they love
Stage
Status: Experiment
Web App
Details
Where it stands
Checked 19 November 2026. The public site loaded with its assistant prompt links, a free skill download link, and a fictional example summary. It asks for no account and says it shows no reviews or user counts yet. I did not download the skill or run it on records.

01Skills

Free skills and prompts.

Skills and prompts I publish for working with AI agents, free and in full. Each lives in its own public GitHub repository: install it, fork it, or suggest a change.

5 free skills and prompts

  • Prompt · MIT

    AgentTeamSetup

    One prompt that sets up a Chief of Agents: it interviews you, proposes three to six role-based bots, and runs them under draft-by-default safety rules.

    Works with: Claude · ChatGPT · Grok

  • Skill set · Free

    AI Lingo + Session Skills

    Skills for AI-assisted thinking, session orchestration and project hygiene, including AI Lingo: mode-switch words such as “first-principles” or “binding constraint” that change how the agent reasons.

    npx skills add sunnypatneedi/skills
  • Starter kit · CC BY-SA 4.0

    claude-starter-kit

    Ready-to-use agents, skills, hooks and MCP configurations for productivity, business and coding in Claude Code and Claude Cowork.

    Works with: Claude Code · Claude Cowork

    /plugin marketplace add sunnypatneedi/claude-starter-kit
  • Skill set · CC BY-SA 4.0

    ClaudeAsk

    A skills framework for Claude Code in which Claude asks you short questions during work and uses the answers to improve your skills’ accuracy and routing.

    Works with: Claude Code

02Selected work

Built at scale.

I spent seven years leading three engineering teams at Apple Maps, building large-scale data and ML systems. Now I invent and prototype AI-native experiences for human learning, collaboration, and dependable agents.

Earlier, I helped build products and intelligent systems at Microsoft and Salesforce.

01 Engineering leadership · Data quality · ML evaluation

Apple Maps

Shipped at scale
Experience overview · schematic
Transit cards & fares
EV routing
3D Maps

Making complex location data dependable for critical product experiences.

7 years · 3 team-supported launches · vendor onboarding: 3 months → 2 weeks.

Challenge
Make complex, continuously changing location data dependable enough to support critical product experiences.
Contribution
Led three engineering teams across Maps experiences, ingestion infrastructure, data quality, and evaluation; partnered across product and engineering to ship complex requirements.
Evidence
Teams contributed to Transit cards and fares, EV routing, and 3D Maps. Ingestion and validation improvements shortened standard vendor onboarding from 3 months to 2 weeks.
What it demonstrates
Technical leadership, systems thinking, cross-functional execution, and engineering judgment at scale.

The Apple Maps schematic is an original conceptual illustration, not a product screenshot. I’m happy to walk through the Apple Maps work in more depth in conversation.

Explore more projects

03How I work

Discover. Make. Verify.

I enjoy the full journey, especially when the problem isn’t well defined and the product doesn’t exist yet.

  1. 01

    Discover what matters

    Understand people, their problems, and what’s newly possible with emerging technology.

    In practice: Rally’s product contract
  2. 02

    Make it real

    Move from ambiguity to a useful prototype. Build, test with people, and refine the experience.

    In practice: Cue’s native apps
  3. 03

    Verify the outcome

    Evaluate behavior, find failure modes, and build appropriate oversight into dependable systems.

    In practice: SayMake’s guardrail evals
Career and role background
Sunny Patneedi

Career

At Apple and Microsoft, I worked on complex systems. As a founder, I bring that experience to hands-on product building: connecting the vision to the technical details, bringing people together and taking ownership when the path is unclear.

  1. SayMake

    Founder & solo builder

    2025 — now

  2. Apple Maps

    Senior engineering manager

    My teams shipped iOS 16 Transit cards and fares, EV routing, and 3D Maps.

    2018 — 2025

  3. Salesforce

    Senior software engineer

    Built production fraud detection with deep learning.

    2015 — 2018

  4. Microsoft (Bing, MSN)

    Senior engineering manager

    Co-led Bing Travel from zero.

    2007 — 2015

Bios and headshots

What I'm looking for

I want to own a consequential AI product, end to end, inside a strong team.

End to end

  1. Customer problem
  2. Technical choices
  3. Launch
  4. Measured results

The mandate

  • Direct access to customers
  • Authority to make product and technical decisions
  • Capable colleagues
  • One outcome I am accountable for

My evidence: my teams shipped Apple Maps features. As a founder, I build products end to end.

Discuss a role

04The problems I want to solve next

Consequential work, end to end.

I’m interested in inventing AI-native products, human-agent collaboration, dependable autonomous systems, and experiences that expand human capability and creativity.

Looking for Staff+/Principal Product Engineering, zero-to-one technical leadership, or Head/Director of AI Product Engineering opportunities. The best fit combines direct customer understanding, meaningful technical authority, hands-on building, excellent collaborators, and ownership of a consequential product outcome.

05For founders and product engineering leaders

Let’s build something important.

If you’re building something where the technology is promising but the right product is still being discovered, I’d love to talk.

I’m especially interested in teams that value invention, technical rigor, thoughtful product design, and meaningful end-to-end ownership.

Discuss an opportunity (opens in a new tab)

A hiring conversation, directly on LinkedIn.