SHELLY X NI

My team and I spent a full research month in Pittsburgh, living together in the same apartment, and working out of the City-County Building. Above is a series of thank you notes we wrote to the many hard-working city staff we met, interviewed and shadowed.

If you're a fire-fighter, paramedic, or any other city staff who responds to emergencies, procurement can be a matter of life or death. When we visited the Emergency Medical Services department, we discovered that life-saving needles were often in short supply, because the one vendor who usually supplied them was not notified about the bidding window for a new medical supplies contract. 

We noticed staff in every department kept their own list, often offline in the form of a private spreadsheet or paper notes like these, of contracts they most often used to buy tools their teams needed. Why? The existing inventory of city contracts was a 500+ row spreadsheet that was difficult to parse, and did not include keywords or line items that city staff recognized. So people keep their own records. This becomes an issue when offline systems don't update when the original spreadsheet does, and staff end up buying items from expired contracts.

We also discovered that the city's 500+ contracts are managed by 2 purchasing staff, using the above spreadsheet. Not only is this highly specialized spreadsheet impossible to understand or analyze by other city staff who need to make purchasing decisions, it often led to details falling through the cracks. A mistake like forgetting to loop in all of the departments involved in renewing a safety equipment contract can result in one department receiving equipment incompatible with previous purchases.

Procurement processes span multiple departments, who don't talk frequently. We brought together stakeholders from all departments involved in the life-cycle of a contract to map out the entire process and identify bottlenecks.

One of the results from the collaborative mapping exercise is this flow chart of different processes city staff choose from when purchasing goods and services.

As a team, we facilitated meetings with major stakeholders to decide on long-term outcomes to pursue during the Fellowship year. We aligned around 4: increasing the number of businesses responding to City contract opportunities, reducing the amount of time purchasing agents spent managing expiring contracts, reducing staff confusion around procurement policies, and reducing the amount of time city staff spent looking up information about existing contracts.

For each outcome area, I led synthesis of our research findings into pain points, needs, and ideas to prototype. 

For each of our 4 outcome areas, I led and documented team conversations about assumptions and open questions on a Trello board. Keeping track of assumptions helped us prioritize what prototypes to build or research conversations to have. 

The first assumption we wanted to test around getting businesses to bid on City work was that providing clear information online, combined with simple outreach (signing up for a mailing list, directly emailing relevant businesses we found on Yelp), would work. 

We asked Citiparks, with whom we'd had a few meetings during our initial research month, if they'd be willing to try a new way of marketing their upcoming cleaning services RFP. We set up a quick github page, created a Mailchimp newsletter, and wrote email templates. 

Our efforts led to 50+ cleaning services signed up to find out about the contract, and 16 times more businesses sending in a proposal than had on previous cleaning contracts!

The success of our pilot site validated our assumption that more businesses would bid on a City contract opportunity if more of them were told about it. Our next step was creating an online resource like our pilot, for all City contract opportunities. I mapped out tasks a business owner would need to be able to do in order to find out and sign up for notifications about City opportunities, starting from how they found out about the site, to how they could share it with their friends.

The City of Pittsburgh and Allegheny County purchase commodities and non-professional services in partnership. They both use NIGP codes to categorize the types of goods and services they buy. These codes are assigned by the NIGP -- the National Institute of Government Purchasing. The category names of each code are often too vague or outdated to understand, so we replaced them with human-friendly language where necessary. 

We conducted user tests with local businesses of all sizes and types. We were especially interested in making sure that they were able to sign up for the right categories for their businesses, and that the value offered by the site was compelling enough for them to complete the sign-up.

After user testing sessions with businesses and city staff, we started to build the site. We changed the sign up flow from categories first to email address input first, from feedback from businesses that an email was a lower barrier to entry. 

We named the site Beacon, and launched the first working version in August 2015 to business communities we'd established relationships with from previous visits.

In addition to building the site, we also needed to plan, design, and write confirmation, reminder, and informational emails. 

A key component of Beacon is staff writing and posting opportunities. Previously, the City did not have a clear process for advertising opportunities. By introducing Beacon, we're able to give staff a simple, single path to spreading the word about opportunities online. We conducted workshops to train staff and answer questions about the process change.

We wrote up a quick reference document to complement our Beacon training workshop.

After internal training, more user testing, and bug fixing, we launched Beacon in September 2015 at a City of Pittsburgh press conference announcing Pittsburgh's greater Roadmap for Inclusive Innovation. 

We followed up our press conference launch with social media flyers like this one.

In the 2 weeks following our launch, a total of 250 Pittsburgh area businesses have signed up for Beacon notifications!

We visited New York City SNAP offices and interviewed folks who were waiting. The first thing I noticed was long lines of people, quietly passing the time on their phones, ignoring the No Cell Phone Use signs in the lobby. Wait times could easily surpass 4 hours.

A study by Blue Ridge Labs confirms what we noticed in the office. 70% of low-income Americans access the internet primarily through their phones.

We hypothesized that meeting applicants where they were — on their smart phones — could significantly reduce the amount of time it took to apply. The time it takes to apply in person can be a barrier for busy parents or full-time workers. 

Though there are many official online applications for SNAP, none can be used effectively on a smartphone. We set out to change that.

SNAP is a federally funded program, which means that states are incentivized to encourage qualified residents to apply for benefits. A major reason people don't is due to perceived and lived experiences of the difficulty of applying.

15 million Americans qualify for SNAP nation-wide, but have not successfully applied. Multiply that by the average monthly benefit, and the result is a huge swath of safety net not reaching people who are food insecure.

We worked with legal aid organizations and benefits experts to edit down the 10+ page paper application to a digestible number of required questions.

Next, I worked closely with Kevin and Wenting, our team's first developers, to quickly prototype a mobile-friendly web version of the application. 

Breaking down required questions into logic flows, and sketching out different types of question screens.

I printed out paper miniatures of each screen, and used the map to plan future features and refine interactions with the team.

We went through 2 rounds of in-person testing on our prototype before launching Easy Food Stamps with a small number of live applicants. Through testing, we learned that applicants trusted Easy Food Stamps more because it was not the official government application.

Once an applicant completes and signs their application, Easy Food Stamps converts it into a PDF, which is then faxed to the right SNAP office for processing. The team follows up by SMS to check on the status of the application and provide additional service.

If applicants are unsure of whether or not they qualify, they can answer a couple quick questions to estimate their eligibility. If they happen to be eligible, they can start an application through Easy Food Stamps with the information they've already entered.

Applicants who don't have easy access to a scanner or fax machine likely do have access to a phone with a camera. Here, they're guided through what documents they'll need to provide, and how to get them.

I led a Kickstarter campaign to sustain the team while we were between funding from Blue Ridge Labs and the next accelerator, Fast Fwd. We raised $11,000+ from 271 backers over 3 weeks.

Propel was accepted to the City of Philadelphia's Fast Fwd accelerator in September of 2014. We started our research in Philadelphia, to understand the Pennsylvania SNAP application process, by visiting benefits offices and recruiting folks for interviews and shadowing sessions. We wanted to know how the application experience differed from what we heard in New York, and we sought to learn more about how Philadelphians on SNAP shopped for food.

We talked to 7 people in-depth, and shadowed 2 of them as they went on a biweekly grocery shopping trip. 

Our research indicated that neighborhood-oriented grocery stores could make great partners for Easy Food Stamps. They could get the word out to shoppers while food was top of mind, and in turn, we could help their customers get more funds for groceries. This is the start of a storyboard for pitching our idea to groceries.

A snippet of the final storyboard we used to explain a Propel pilot in the context of a local grocer. The pitch was successful, and we ran our first in-store pilot in December of 2014.

Post-card sized flyers we handed out during our grocery store sign-up pilot.

Through Fast Fwd, we created buy-in from government staff at the City of Philadelphia and the State of Pennsylvania. This allowed us to submit and keep tabs on applications we sent in more easily. 

The next phase of work is prototyping grocery store interventions, with the goal of screening people for SNAP eligibility and marketing Easy Food Stamps in a place where food is top-of-mind. 

We're currently piloting an in-store marketing project with a Philadelphia grocer. 

Final version of print guides, used here by Reconnect employees.

Reconnect Cafe is a special place in Bed-Stuy, Brooklyn. It's a non-profit cafe that employs young men from the neighborhood seeking more stability in their lives. It offers job training, a fair wage, and connections to job placement and tutoring programs. 

I started volunteering there since I wanted my thesis project to have immediate impact in my local community.

I spent the first few months providing pro bono design work and hanging out with the cafe crew. Together, we brainstormed and designed engaging activities for their Grand Opening. Shown here is a vegetable-powered beat mixing station, powered by a Makey-Makey and ingredients the crew uses in their innovative juice recipes.

Five months into volunteering, I helped one of the employees go to the Social Security Office to replace a lost social security card. What I thought would be an hour errand blossomed into a 5-hour ordeal.

T.I. needed an ID to apply for a job training program. Between moves, he had lost his passport and EBT card. He sent off his social security card to get his EBT card replaced and it had gotten lost in the mail. His attempt to use a valid ID, his high school diploma, was rejected, because he did not pair it with the right supplementary ID.

According to another Reconnect volunteer, Kia, difficulties with ID was a recurring issue at Reconnect.

I asked myself -- what if I focused my thesis work on designing a service to help people navigate government application processes, to get basic IDs?

Combining official application instructions, print and online, with feedback from social workers, I started mapping out all the necessary steps.

Having mapped out the abysmal current experience, I used Freytag's Triangle as a framework for writing out an ideal ID application experience.

Using color-coded steps, I started putting together a "choose your own adventure" style walkthrough for social security card replacement applicants.

I tested the books at Reconnect and the NY Common Pantry, a food bank in East Harlem that also provides services to homeless clients.

Books were onerous to leaf through -- user testers wanted quick access to information specific to their situation. I started prototyping how-to-guides as pamphlets instead.

There was sufficient interest from social workers at food banks and homeless shelters, and other volunteers at Reconnect, for a facilitated guide that workers could use on desktops, that I prototyped a web experience.

A mobile-friendly website would be the ideal digital solution, since many low-income New Yorkers have access to smartphones.

I asked the 20-odd probation clients and handful of probation officers I worked with in the South Bronx to rate the usefulness of each type of guide, as well as the usefulness of my system against existing sources of information. I received a lot of helpful positive feedback.

Versions of the pamphlet and mobile website, as well as a sample text from an SMS version of the service.

A 3 minute user journey video.

We prototyped Nom by projecting a website onto a kitchen counter with an ordinary projector. Nom was 'driven' using a computer and a mouse to great effect.

The home screen for Nom prompts the cook through browsing recipes, or seeing what's on hand in the kitchen.

From our interviews with novice and expert chefs, we recognized that a key difference between the two groups was the expert's ability to define and combine flavors. Nom helps novice chefs by suggesting dishes that go together, and opens up new combinations to experts as well.

Not only did we explore basic flows and button placement with wireframes, we also needed to test how colors showed up on non-traditional surfaces like kitchen counter wood and food.

We learned that a neutral beige color appeared nearly translucent on natural wood surfaces.

Rigging the projector securely was key to setting up a testing space.

We asked novice chefs to user test recipes from cookbooks, by talking aloud as they planned how to prepare them. 

We learned that it was most useful to structure instructions by calling out the action first, and then the ingredients involved in that action. 

We also linked to existing instructional videos from the actions, since many actions, like basic knife skills, are transferrable across recipes.

Wireframe of a recipe

Wireframe of a shopping list for a trio of recipes.

Wireframe of a specific interaction for cocktail shakers.

Connecting Nom features to common issues that arose from research with novice cooks.

Good meal prep leads to great meals. This map shows how Nom connects together existing services, like a smart kitchen inventory and grocery delivery, to improve the entire meal experience.

Boyshorts available for sale at The League of Ladies store.

Frida's superpower is the power of self-expression. What I admire most of her legacy is her fearlessness in depicting her perception of the world.

Amelia's superpower is the power of flight.

Marie's superpower is the power of intellect. Each character came with an accompanying zine, which led the reader through a short activity related to the character's superpower.

Watch the 2 minute video!

Long before storyboarding the video or writing the content for the campaign, I was actively engaging a Tumblr community and building a mailing list of interested feminists.

The League started as a playful idea: could I sell feminist underwear to strangers on the internet? 

I began, as with all projects, with conversations. I interviewed women I admired about what feminist meant to them, and what apparel, if any, made them feel more confident in their day-to-day life.

I heard great stories about Wonder Woman Underoos, stealing superhero boxers from the boys' section, women who had chose underwear based on how they wanted to feel that day, and had important conversations about the legacy of activism in New York. 

In a conversation with a mentor, Jeremy Canfield, the dots between honoring past feminist legacies and the clear interest in superhero underwear connected. And thus, the League was born.

I posted preliminary sketches and thoughts to The League of Ladies blog, because I believe in sharing work early and often. Presenting unfinished work to the world gave fans the feeling that they could jump in and contribute.

After successfully raising Kickstarter funds, I switched into full prototyping mode. I worked closely with my screen printer to try different placements of superhero faces and symbols, and surveyed Kickstarter supporters for feedback.

A huge part of Kickstarter is being along for the ride, and I wanted to honor that.

Variations of bands used to enclose League of Ladies packages.

In the fall of 2014, we were honored to collaborate with an amazing women's underwear start-up, Dear Kate, on a line of League of Ladies underwear for their collection. 

We added a 4th superhero, Harriet, inspired by Harriet Tubman.

From the Dear Kate product catalog.

From the Dear Kate product catalog.

From the Dear Kate product catalog.

From the Dear Kate product catalog.

I worked with the 2012 Feast Challenge team to research and synthesize the needs of aging New Yorkers into actionable hackathon challenges.

Gaia orain and I interviewed 40+ suburban and urban seniors on the future of aging in place for a medical alert company.

Designing a diabetes management program for the unique needs of East Harlem residents for City Health Works. City Health Works is a social enterprise that trains peer coaches, who are trusted residents of the neighborhood, to teach their neighbors healthy cooking and exercise habits.