Launch HN: Vela (YC W26) – AI for complex scheduling
Vela launches an AI-powered agent to tackle the notoriously complex, multi-party, multi-channel scheduling problem, aiming to eliminate the endless email chains and manual coordination. The Hacker News community dives into the true definition of
The Lowdown
Vela introduces an AI agent designed to automate complex, multi-party, multi-channel scheduling, aiming to eliminate the back-and-forth typically associated with coordinating meetings. The founders, Gobhanu and Saatvik, frame scheduling as a "constraint satisfaction problem disguised as email," highlighting the difficulty of handling unstructured natural language, evolving constraints, and subtle social dynamics across various communication platforms. They envision a future where scheduling "just happens," freeing up significant time for professionals.
- Core Functionality: Vela integrates seamlessly across diverse communication channels, including email, SMS, WhatsApp, Slack, and Applicant Tracking Systems (ATS). It reads context, checks calendars, proposes times, handles follow-ups, and manages rebookings automatically.
- Real-world Problem Solving: The service targets industries like staffing, where coordinators juggle hundreds of candidate-client interviews across different time zones and communication methods, often encountering cascading reschedules and mixed-channel responses. Vela claims to have solved long-standing scheduling woes for such clients in minutes.
- Data-Driven Approach: Acknowledging the "data problem" as the hardest part, Vela has built unique behavioral datasets capturing response latency by role, channel preferences, follow-up timing, and optimal option presentation to avoid decision paralysis.
- Technical Challenges: The system addresses significant technical hurdles such as maintaining state across channels (unifying identity, merging context), managing temporal natural language understanding ("next Friday"), and discerning when to infer information versus asking for clarification.
- Current Status: Vela is live with paying enterprise customers, continuously encountering and adapting to new edge cases. Demos and case studies are available on their website.
The founders are actively seeking feedback from experts in multi-agent coordination, conversational AI across channels, and constraint satisfaction in complex real-world environments, eager to discuss their approach and challenges with the Hacker News community.
The Gossip
Complexity Critiques & Clarifications
Initial comments questioned Vela's claim of "complex scheduling," comparing it to simpler tools like Doodle or suggesting their case studies didn't fully illustrate the intricate problems. Critics argued that basic scheduling involves mere preferences, while true complexity lies in multi-pending invites, urgency, and relationship importance. Vela's co-founder responded, clarifying that simplified case studies mask underlying multi-party dependency resolution and concurrent problem-solving capabilities, which are central to their value proposition.
Market Focus & Niche Navigation
Commenters offered strategic advice, strongly recommending Vela "niche hard" into specific industries like logistics or executive search, rather than attempting a horizontal solution, citing the success of companies that focused on messy, industry-specific edge cases. Discussions also explored potential B2B applications, such as scheduling surgeries, while some individual users expressed interest in a B2C offering, to which the founders confirmed a primary B2B focus.
Algorithmic Architecture & AI Accolades
The technical community chimed in on the underlying AI and optimization challenges. Several users shared their own experiences with constraint satisfaction problems (CSPs) and optimization solvers, acknowledging scheduling as an NP-hard problem. While some praised the agent-based approach for handling real-world rescheduling loops, one AI researcher pointed out that the "scheduling solved" slogan might be misconstrued by those familiar with advanced AI optimization techniques like GNNs and RL for complex industrial problems. Vela's team acknowledged this feedback, emphasizing the role of LLMs as a tractable layer above the CSP modeling and state management.