Sudo Make World 2026 — Open-Source Tools for Social Good

Twenty-six teams had 72 hours to build open-source tools for social good. sudo make world 2026 ran February 27 through March 2, challenging participants to create software that addresses real-world problems in education, climate, healthcare, civic engagement, and community infrastructure. Projects were scored across Impact & Vision (35%), Technical Execution (25%), Innovation (20%), Usability (15%), and Presentation (5%) by 38 engineers across three evaluation batches.

✯ Grand Prize Winners

• 1st Place — Witness (Team: Witness). A human rights documentation platform that transforms raw field testimony into ICC-standard evidentiary memos with chain-of-custody protocol. Whisper large-v3 transcription, Mistral Large legal annotation, cross-referencing against ICC, UN, ACLED, Amnesty International, and Human Rights Watch databases. Score: 4.612/5.00. Total winnings: $1,400 (1st Place + Best Technical Execution + Best Innovation + Best Usability + Best Presentation). "The standout submission of this hackathon, and one of the most purposeful civic tech projects we've seen at this level." — Irina Titova
• 2nd Place — Scam_BaitAI (Team: NinjaCodes). An AI-powered anti-scam platform that actively engages fraudsters with convincing personas, extracting intelligence while protecting real victims. LangGraph agent orchestration, hybrid ML detection (TF-IDF + SVM), full voice pipeline (Twilio, Deepgram, ElevenLabs), 30+ concurrent conversations. Score: 4.587/5.00. Total winnings: $400 (2nd Place + Best Impact & Vision). "Every minute the bot holds a scammer's attention is a minute they're not targeting a real victim." — Irina Titova
• 3rd Place — NextPlate (Team: BetterWorld). A food waste reduction platform connecting restaurants, NGOs, and customers through Ghost Meals, Recipe Alchemist, and WRAP-based carbon tracking. Multi-key Gemini pool with exponential backoff, PWA, i18n, multi-role authentication. Score: 4.462/5.00. Prize: $200. "The level of architectural maturity shows a professional-grade approach to a complex logistics problem." — Venkata Revunuru

✯ Category Excellence

• Best Impact & Vision: Scam_BaitAI (NinjaCodes) — Highest impact score at 4.75/5.00.
• Best Technical Execution: Witness (Team Witness) — Highest technical score at 4.75/5.00.
• Best Innovation: Witness (Team Witness) — Innovation score at 4.50/5.00.
• Best Usability: Witness (Team Witness) — Usability score at 4.75/5.00.
• Best Presentation: Witness (Team Witness) — Presentation score at 4.75/5.00.

✯ Excellence Tier

• Foresight (Decision Dynamo) — Structured decision-making platform with scenario projections, color-coded timelines, and irreversibility flags. Score: 4.375/5.00
• LifeVault (DINooo) — Offline-first secure document vault for displaced populations. Local-only storage for refugees who cannot trust cloud infrastructure. Score: 4.288/5.00
• AfterWord (Hanuman Force) — Digital estate management helping families discover and close accounts of deceased loved ones. Guardian verification and AI-powered discovery. Score: 4.125/5.00
• Offline-Learn (lossy bird) — Education platform designed to function without internet connectivity, targeting 2.6 billion people without reliable access. Score: 3.987/5.00

✯ Standout Innovations

• ArchAI (WOLF) — Multi-layer engineering intelligence system with analysis context lazy-cache pattern and workflow projection engine. Score: 3.975/5.00
• CodeJam (The Experimentalists) — Collaborative coding platform fostering developer community through structured challenges and peer review. Score: 3.937/5.00
• ReliefNet-AI (CrisisCommander) — Climate-aware disaster response platform with NLP incident classification and explainable AI workflows. Score: 3.925/5.00
• Refugee Ready (Team Dua) — Multilingual refugee assistance with in-browser OCR, "first 72 hours" survival features, and offline accessibility. Score: 3.781/5.00
• SELF-HEALING NODE (Lone Wolf) — Autonomous code repair agent: analyze repo, fork, fix, and submit PR without human intervention. Cyberpunk UI. Score: 3.400/5.00

✯ Evaluation Panel

Projects were evaluated by 38 engineers across three batches, including professionals from Okta, Google, Cisco, AWS, and other technology companies.