Kilo Co‑Found Special Track: AI Agents for Everyday Well‑Being

Unfair Weekend has a new special track. We are partnering with Kilo Co-Found to look for AI agents that help people live healthier, calmer, more connected and more capable lives. The team whose project fits the brief best wins a Founders’ Dinner with the Kilo Co-Found programme leaders.

Founders’ Dinner with Kilo Co-Found: the selected team joins the programme leaders for dinner, product feedback and an honest conversation about growing their idea.

Who is Kilo Co-Found?

Kilo is a Lithuanian venture builder that launches and scales consumer brands. Through its Co-Found programme, Kilo backs builders with capital, operational support and access to its in-house platform for data, AI, engineering, growth and creative work. Co-founders validate an idea within about six months and receive equity once the business proves it can stand on its own.

Kilo focuses on subscription and direct-to-consumer products with a real repeat habit: wellness, active lifestyle, mental health and digital utilities. That is why this track exists. The team is looking for the kind of product it would want to build next.

What the prize is, and why it matters

The whole selected team joins the Kilo Co-Found leaders for dinner. Expect direct product feedback, mentoring and a candid conversation about what it would take to turn the prototype into a company. If there is a strong match between the team and the programme, the conversation can continue after the event.

The track sits alongside the main prizes. Build what you would build anyway; if it falls within the brief below, you are in the running for the dinner too.

The brief: agents that own a recurring job

Kilo is not looking for another dashboard, content library, chatbot or habit tracker. It wants AI-first consumer products that take responsibility for an ongoing job. The agent should understand context, remember the user, make decisions, coordinate actions and improve with repeated use.

The strongest products share a retention loop:

1. The user connects data, preferences, goals or relationships.
2. The agent builds a persistent understanding of that person.
3. It takes useful actions or makes decisions on their behalf.
4. The user returns because the agent keeps getting more accurate, more helpful and harder to replace.

Start with one painful, frequent use case. A broader personal operating system can come later.

Five directions to build in

1. A personal health agent that actually does things

People already have health data in wearables, medical records, lab results and food logs. What they lack is help deciding what matters and what to do next. Build an agent for ordinary people, not biohackers. It connects a few useful sources, such as Apple Health, Google Health Connect, one wearable or uploaded lab results. It learns the user’s goals, routines and constraints, then turns a small number of recommendations into concrete actions.

The difference is agency. A weak version says, “Your sleep was poor.” A strong version notices three bad nights before an early presentation, suggests a lighter workout and an earlier wind-down, and offers to reorganise the evening. It explains uncertainty and keeps wellness guidance clearly separate from diagnosis.

2. An AI household OS for family health

In many families, one person carries the medications, appointments, records, children’s symptoms and paperwork for everyone. Build an AI caregiver that coordinates that workload while respecting each person’s privacy. It keeps separate profiles and permissions, accepts information through conversation, voice, photos and forwarded documents, and turns instructions and prescriptions into tasks and follow-ups without flooding anyone with notifications.

This is not a digital filing cabinet. It should do the coordination work. Good starting points include parents managing children’s health, families living with a chronic condition, or adult children caring for ageing parents.

3. A midlife performance agent for women

Menopause apps tend to track symptoms or deliver content. Fitness apps focus on training, and longevity apps focus on biomarkers. Build the agent that connects them. It understands a woman’s changing symptoms, sleep, training, work and caregiving load, then adapts daily decisions about strength, recovery, nutrition and energy.

It should answer practical questions: train hard or recover today? Why has my energy changed this week? What should I raise with my doctor? Over months it learns which changes actually help her, and shows progress in strength, sleep, energy and confidence rather than app engagement.

4. An agent for nervous-system regulation

Meditation and breathwork apps are mostly libraries: pick a session, play it, leave. Build an agent that recognises the moment and starts the right intervention. It learns a person’s stress patterns from lightweight signals such as self-report, sleep, heart rate, calendar context and time of day. It spots recurring situations like pre-meeting anxiety, the post-work crash or trouble falling asleep.

It begins by recommending short protocols, then earns permission to act on its own. Picture this: “Your meeting starts in 20 minutes and your heart rate is up. Start a three-minute downshift now. I’ll mute non-essential notifications and check in afterwards.” Measure success by changes in how people feel, not minutes listened.

5. Open category: agentic consumer products

Have your own idea? Kilo welcomes ambitious AI-first consumer products in health and wellness, fitness and recovery, family and ageing, beauty and self-care, personal development, education, travel, consumer finance, entertainment and adjacent categories. The product must be more than an AI feature. Think of a tutor that notices where a learner gets stuck and adapts the practice, or a self-care advisor that learns what works and manages the routine.

For an open-category project, be ready to explain:

• What recurring job the agent owns, and why people need it often.
• What it can do that a normal app cannot.
• What memory it builds and which actions it takes.
• How the value grows over time until it becomes a habit or an indispensable system.

What Kilo is not looking for

• Generic AI chatbots with health branding.
• Passive dashboards for wearable data.
• Static content libraries with an AI search box.
• Habit trackers that leave all the work to the user.
• “Everything for everyone” health super-apps.
• Wellness advice without measurable outcomes.
• Products whose only retention mechanism is push notifications.
• Medical diagnosis or treatment claims without proper clinical safeguards.

What a strong prototype shows

You have four days, not a year. The Kilo Co-Found team will look for a focused proof that points to a bigger product. A strong demo shows:

• A specific user with a recurring, painful job.
• An agent that acts, not one that only answers questions.
• Persistent memory that visibly improves personalisation.
• A clear permission and safety model: what the agent may do alone, and what it asks first.
• At least one integration or realistic data source.
• A measurable outcome beyond engagement.
• A credible path to revenue, such as a subscription, family plan or transactions.
• A retention loop that explains why the product is more useful after 30, 90 and 365 days.

Health data is sensitive. Use fictional or consented sample data during the hackathon, and say clearly in your demo which parts are real, which use sample data and which are unfinished.

How to take part

There is no separate application. Apply to Unfair Weekend, join a team and build a project within one of the five directions above. Projects that fit the brief are eligible for the Founders’ Dinner, and the Kilo Co-Found leaders choose the team.

Not sure how to shrink a big idea into a weekend-sized demo? Read the AI hackathon ideas guide and how to prepare a convincing hackathon demo.