1. Background & Problem
Describe:
- The real-world pain point
- Who the users are
- Why the problem matters
- What triggers the need for this solution
(Example: “Ovulation detection on BBT curves was slow, inaccurate, and required user intervention…")
2. My Role
Clear, factual statements of your responsibilities:
- Led algorithm design for…
- Owned end-to-end product strategy for…
- Conducted model research on…
- Collaborated with hardware / app / data teams on…
(This section is extremely important for hiring managers.)
3. Technical / Product Solution
Break down the solution into multiple sub-sections depending on project type.
3.1 System Overview
A high-level narrative of how everything works together.
3.2 Architecture Diagram (optional)
Insert an image:
3.3 Algorithms / Models
Explain:
- Methodology
- Key equations / statistical approaches
- CNN / LSTM / transformer usage
- Signal processing pipeline
- Calibration or preprocessing steps
- Training strategies
3.4 Data Pipeline
Describe data:
- scale
- preprocessing
- labeling
- storage
- syncing
- augmentation
3.5 Product Decisions
If applicable:
- UX choices
- trade-offs
- onboarding flows
- user GPS/sensor constraints
- business model considerations
You can delete irrelevant sections depending on the project.
4. Challenges & Trade-offs
Discuss constraints such as:
- hardware limitations (sensor noise, sampling rate, battery)
- model size vs accuracy trade-off
- deployment constraints (mobile CPU/GPU)
- user behavior diversity
- ambiguous biosignals (HRV, BBT, sleep)
This section shows engineering maturity.
5. Results & Impact
Use numbers if possible:
- accuracy
- time reduction
- performance improvement (%)
- number of users impacted
- growth achieved
6. What I Learned
A concise reflection:
- insights
- mistakes
- how you would improve it
- how it influenced your thinking on algorithms/product
This is critical for graduate school essays & interviews.
7. References
Possible elements:
Screenshots
Diagrams
Videos
If stored on Google Drive or YouTube:
Other References (Optional)
Links to pdfs, papers, books, engineering docs, or open-source libraries used.