Puente

MS&E 265 · Stanford · 2026

Stanford’s row house chefs cook for 55 to 75 people a day, nine months a year. But when the school year ends, the paycheck stops.

Puente is a seasonal job matching platform for those chefs. Jobs arrive as a WhatsApp message and the chef replies YES or NO, so there is no app to download, no login, and nothing to search. A Stanford employer-verified return guarantee means accepting summer work never risks the fall position.

Shape
Marketplace · Two-sided
Channel
WhatsApp-native
Year
2026
Fig. 1 · Home screen

Abstract

01After interviewing row house chefs and the CEO of Student Organized Services (their employer), we realized the real problem wasn't the shortage of summer jobs, but a job search prcocess that assumes tech literacy, English, and no risk to the fall position.
02We put 15 ideas on a feasibility × impact matrix, then prototyped 6 and tested them with chefs to solve the root issue of summer employment.
03Jobs arrive over WhatsApp instead of an app, the reply is YES or NO instead of an application, and an employer-verified return guarantee means taking summer work does not cost them the fall job.
01Problem SpaceSeasonal income gap

Who the customer is

Row house chefs are contracted food service workers in Stanford’s self-operated and Greek housing system, serving roughly 27 houses through Student Organized Services (SOS). They work about 34 weeks a year, late September to mid-June, with no employment over summer, winter break, Thanksgiving, or spring break. Each chef plans, preps, and serves two daily weekday meals for 55 to 75 people, handling menu planning, ordering, portioning, storage, dietary accommodations, and last-minute guest counts.

Most started in entry-level roles like dishwashing and worked their way up. Average tenure is ten years, so they build relationship with the students, houses, and the University

Then the year ends

In-season wages are $34.39 an hour. Unemployment during furlough runs roughly $400–600 a month, which is functionally meaningless against Bay Area costs.

Finding a job over the summer is never consistent, it’s one of the biggest struggles we have.

Joel, Pi Beta Phi chef

Three barriers to Summer Employment

01

The loyalty trap. Chefs hesitate to take summer work because doing so may cost them their fall spot, and there is no return guarantee on either side. Claudia turned down summer work because the employer wanted her past September, and she did not want to give up her position.

02

The search itself. The process of finding summer jobs has no structure and no support. Joel: “Sometimes there are summers where there is nothing, and I have to resort to construction jobs; not related to cooking at all.”

03

Tech and language access. Digital job boards assume tech literacy. Per Nick, the SOS CEO, staff “struggle with password management and many have over 5,000 unread emails.” Most chefs have English as a second language, which makes the search even more complicated.

Sometimes it is difficult to communicate with students because I don’t know English very well, especially if they’re talking fast. That’s why I always end up speaking Spanish with the kids who know even a little bit of it.

Ana María, Kappa Sigma chef

Sizing it

$764M

TAM: ~849k US seasonal and contract food service workers

$49M

SAM: institutional food service, K-12 and universities

$121.5K

TM: full-adoption ARR ceiling across ~135 Stanford row house workers

We assumed a 10% take rate paid by the employer, well below the staffing industry’s typical 25%+ and on-demand competitors like Wonolo at ~45% since Puente is not a full staffing agency. Chefs keep their entire wage.

02Market Research5 competitor classes · 12 constraints

Existing Gaps in Market

Every existing option either targets the wrong user or is inaccessible because of language, tech, or design.

General job platforms

Indeed · ZipRecruiter · Craigslist

Large inventory, free, broad Bay Area coverage.

Key difference

Built for tech-literate users who manage accounts, passwords, and inboxes. For a chef with 5,000+ unread emails and English as a second language, the interface poses as barrier to entry.

Gig apps

TaskRabbit · Instawork · Shiftgig

On-demand flexible shifts that could theoretically fill the gap.

Key difference

Built for urban gig workers who already live on these apps. Signing up assumes the same fluency, and ▎ the relationship ends with the shift, so there is no community around it and nobody to vouch for what a ▎ chef can do.

Staffing and temp agencies

Local culinary staffing firms

Human intermediaries who can bridge some language and literacy gaps.

Key difference

They take a significant cut of already thin seasonal earnings, and placement is passive and anonymous.

Word of mouth

SOS informal networks

Personal introductions place chefs successfully because real trust is already there.

Key difference

Entirely dependent on one administrator’s bandwidth and personal network. Not scalable across 27 houses, inconsistently available, and undocumented.

Government and nonprofit programs

EDD · workforce development nonprofits

Free, ESL-friendly, designed for low-income workers, with unemployment insurance as a floor.

Key difference

A floor of $400–600/month does not go far against Bay Area rent. Each chef is expected to work through the ▎ bureaucracy alone, and they can't leverage their existing relationships.

Across all of these solutions, what was missing was a low-friction way to find seasonal work quickly, without tech fluency, without threatening the fall position, and without starting from zero.

Word of mouth was a promising lead that we could develope. The chefs can leverage surrounding community, meaning residents, alumni, local families and nearby businesses to vouch for them at scale.

Constraints we wrote before we designed anything

Desirability

  • No account creation or management.
  • Must work in Spanish.
  • No cold-applying to strangers.

Feasibility

  • No digital literacy beyond basic smartphone use.
  • Return guarantee issued by SOS, not the platform.
  • Cannot depend on reliable computer access.
  • Cannot require food handling permits or commercial kitchen access.

Viability

  • Scoped to job discovery only at launch.
  • Chefs do not build their own profiles.
  • Free to join for chefs.
  • Cannot depend on a large employer pool at launch.
03Diverge15 ideas → 6 prototypes

Crazy 8s, then a matrix

We generated fifteen concepts, then plotted all of them on a 4×4 matrix across feasibility and impact. Six landed in the top-right quadrant (realistically buildable and most likely to move the needle) and we built out those 6 prototypes.

All fifteen ideas plotted on feasibility × impact. The six starred concepts in the top-right quadrant went forward to prototyping.

The six we built and tested

Each prototype went in front of chefs, less to find a favorite than to find out which layer of the problem each one actually touched.

Cocina English · Bilingual kitchen vocabulary and conversation lessons. Needs no employer participation, but addresses communication rather than job discovery.
Bay Area Chef Match · Employer-facing posting with instant matching. Removes nearly all friction from the chef, but the demand side was unvalidated.
Voz Trabajo · Spanish voice assistant for resumes and interviews. Voice removes the reading barrier, but a resume matters only after a job is found.
Summer Meal Subscription · Chefs list a menu and cook for subscribing families. Predictable recurring income, but assumes kitchen access and permits chefs don’t have.
Alumni Chef Pipeline · Turns the alumni network into a hiring pipeline. Trust is pre-built, but it only works for chefs who already have an alumni relationship.
Puente (v1) · WhatsApp job matching with coordinator-built profiles and an SOS return guarantee. Solved discovery directly, but leaned entirely on one human coordinator.

What testing actually returned

Claudia responded most strongly to Cocina English, connecting it immediately to colleagues who struggle with utensil names and allergen vocabulary. Joel was most excited by Voz Trabajo, calling resumes a weak point and suggesting a feedback loop so chefs who don’t get a job learn what was missing.

But when Claudia evaluated Voz Trabajo, she mentioned the key reason we ended up rejecting that idea:

The hardest part isn’t writing a resume, it’s finding the right opportunity in the first place.

Claudia, row house chef, on the Spanish resume builder
04Converge6 tested → 1 chosen

Testing surfaced three strong contenders: Cocina English, Voz Trabajo, and Puente. The two that people liked most were the two that solved the root issue of finding a job. English proficiency and resume building only matter once a chef has found a job.

Why each runner-up lost

01

Cocina English and Voz Trabajo both help a chef perform once hired.

02

Alumni Chef Pipeline satisfied the no-cold-outreach constraint and got a warm response, but it only works for chefs who already have an alumni relationship, which excludes a significant portion of the user base. Alumni would only hire chefs they personally knew.

03

Summer Meal Subscription violated a key constraint. Claudia had no commercial kitchen access and couldn’t meet food quality standards without one. Permits and liability had no clear mitigation path.

04

Bay Area Chef Match Employer facing rather than focusing on the chefs.

The constraint we broke on purpose

Our original viability constraint said chefs do not build their own profiles, but relying entirely on one human coordinator for all the chefs does not scale. So we adapted our constraint. No, onboarding is self-guided by default, and Nick at SOS agreed to build profiles on behalf of chefs with lower tech proficiency.

Why the demand side was defensible

Nick (SOS CEO) confirmed that fall employment has been protected for chefs who take summer work for as long as the contractor system has existed. Private chef demand in California is rising: $38 to $71 an hour in San Francisco, San Jose, and Atherton, with the global personal chef services market projected to grow at 6.7% CAGR through 2030.

05The Product5 screens · one loop

Puente is a seasonal job matching platform for Stanford row house staff. The core loop is deliberately small:

01

A chef builds a profile, or the employer builds it for them.

02

Job opportunities arrive as a WhatsApp message.

03

The chef replies YES or NO.

04

An employer-verified return guarantee means accepting summer work never risks the fall position.

Home

  • “Your fall job is always protected” sits above the fold. This directly addresses the loyalty trap. Claudia turned down summer work not because jobs were unavailable but because accepting felt risky.
  • The stats row uses real interview numbers. $0 to join, a 3-month income gap, $38 average hourly rate.
  • The EN/ES toggle is in the nav. To help our ESL users navigate the platform better

Find Work

Shows open roles and gives chefs a way to reach out, with the emphasis on work arriving rather than being hunted.

  • WhatsApp match banner. Inbound matches arrive as live notifications with YES/NO buttons.
  • “Spanish OK” filter. Instantly narrows to employers where language is not a barrier. Claudia specifically mentioned this as a bottleneck in her interviews
  • Cards expand in place. Job details open inline instead of on a new screen, keeping the tech literacy requirements minimal
  • Follow-through happens in WhatsApp. Instead of keeping track of another platform, we meet the chefs where they already are interacting.

My Profile

Presents the chef’s qualifications to employers without requiring the chef to write anything.

  • “Verified by Stanford SOS: Fall position protected.”
  • Skills as tags, not prose. Translates experience without requiring a detailed resume. Claudia called a pre-built profile a major relief.
  • Availability as structured data. Dates, hours, and travel radius give employers what they need with zero writing burden on the chef.

How It Works

  • “No app download. No password. We do the hard part.”
  • Step 1: The conversation can be in Spanish. The entry point to the product requires zero technology from the chef.
  • Step 3: Affirms real WhatsApp exchange.
  • Step 5: the return guarantee.

For Employers

The demand side: where families and businesses post jobs, browse chef profiles, and request a chef.

  • Skill tags over resumes. “60+ daily service,” “dietary accommodations,” “bilingual” translate kitchen experience into language employers actually understand.
  • Seasonal acknowledgment callout. Tells employers upfront that all chefs return in September.
  • Free to post. Zero cost removes the biggest barrier to first-time demand-side adoption
  • Browse before posting. Employers discover chefs by proximity, availability, and skills before committing, matching how trust-based hiring actually works.
06Next StepsRoadmap · 3 themes

The Numbers

At a 10% take rate on a $112 average booking, 25 bookings a summer, and an estimated three-summer employer relationship, LTV lands at $714 and the maximum allowable CAC at $238 on a 3:1 benchmark. Four acquisition channels (Bay Area Stanford alumni groups, Nextdoor, a structured referral credit, and Berkeley Parents Network) model out between $2.08 and $28.57. Even the most expensive returns 25:1, which says the constraint on this business is supply-side trust, not acquisition spend.

$714

LTV per employer, over an estimated three summers

$238

Maximum allowable CAC at the 3:1 benchmark

$2–29

Modeled CAC range across four channels

Where it stands

Puente came out of MS&E 265 as a validated concept with a working prototype and a launch model, not a deployed product. I’m still working on it.

The roadmap

Theme 1

Getting both sides onboarded

1

Two paths for Chef onboarding

Chefs with enough English fill out availability, skills, and specialties themselves, the rest answer the same questions through a Spanish voice agent or ask SOS for verification.

Impact HighCost Low
2

Employer onboarding

Families and small employers enter dates, hours, meals per week, dietary needs, location, and budget, then add payment details.

Impact HighCost Medium
3

Match-readiness score

An internal score from profile completeness, verification, and language level.

Impact HighCost Low

Theme 2

Trust and reliability after the match

1

Escrow payment on Stripe Connect

Employers authorize payment at booking, Puente holds it, and it releases when the job is marked complete. This is also where the 10% take rate is collected.

Impact HighCost Medium
2

Booking-day operations

A small Puente team, separate from SOS, confirms the chef has the address, handles a no-show employer, and records that the job happened so escrow can release.

Impact MediumCost Medium-High

Theme 3

Job readiness once a chef is placed

1

Language practice tab

Kitchen vocabulary, phrases used during service and prep, and scenario dialogues in Spanish to English.

Impact HighCost Medium
2

Event prep generator

A chef enters an upcoming event and gets prep for it, like menu vocabulary, service timing, and client scripts.

Impact MediumCost Medium
3

Workplace norms guide

A browsable reference for the unwritten rules of different kitchens and teams.

Impact MediumCost Low