Team
HeyDev
This team is at maximum capacity.
Project Concept
A tool for Developer Relations & Dev Advocates that helps them move faster, stay in sync with their communities, and scale their impact. Saves time manually generating every demo, doc, or insight.
• Auto-generated blog posts and tutorials from changelogs
• GitHub + Discord + X sentiment analysis
• Code sample scaffolds in any framework
• Internal feedback summaries, ready to ship to Product
Entry
Status: Submitted
Last saved: May 18 at 2:56 PM PDT
Team Roster (team is at max capacity)
Message board not available for this team yet.
Lana Zumbrunn Team Lead RSVP Approved
Cofounder & CTO at LevelUp Economy
Project idea
PM
Planning & Documentation
Frontend design and development
CopilotKit integration
Delivering technical education products and scaling them to onboard global talent to AI and ML careers.
Love to build custom solutions with clients.
Background in startup ecosystem building and strategy consulting.
Started an emerging tech community, Rooftop Global, for women to learn about and invest in emerging tech including AI and blockchain.
Run outdoors, hike, am a serious foodie.
Onboarding others to applied AI and ML systems
Software factories
Building AI applications
Automation
Web3 projects
Designing and teaching AI/ML bootcamp curriculum in the Middle East (Jordan)
Built a "curriculum factory" - a multi-agent pipeline that generates lectures/slide decks, applied labs with GitHub Classroom assignments, drills, and instructional resources spanning from demos to grading rubrics, all from a single spec.
Education products
Working with clients to implement AI products their workflows
Integrations, and AI features for the Rooftop community, rooftop.global
Harshdeep Gupta RSVP Approved
Software Engineer II at Microsoft
Idea and vision collaboration with team,
Copilot kit, crew ai flows for github to doc generation
Debugging and Developing, mostly debugging
I am in tech because I like to build stuff. Working at a big tech company working as a systems engineer to build distributed computing systems.
Outside of work I dance and read books, and passionate about real estate and investing in it on the side.
Learn to build saas apps, and experimenting with AI coding agents
Looking for cofounders and an community to connect with and call home
A website to help brands discover content creators. Using Claude to code in frameworks where I have no prior experience.
And a few others , including an app to help newly weds to sort through wedding pictures, and latest one being a nutrition AI agent, powered by langchain.
Hari Panjwani RSVP Approved
Software Engineer at Amazon
Problem: Created posts / editing livestreams video and posting on social media feature.
Livestream video length reduction, removing quiet parts automatically.
Attempted creating blog post out of video, taking out photos from video, to give glimpse of event for post.
Used agent crewai workflow to accomplish the above.
Hari Vilas Panjwani is a software engineer at Amazon Web Services (AWS) specializing in AWS serverless API gateways. With approximately 4 years of experience, he focuses on developing scalable distributed systems and implementing innovative solutions in software engineering and machine learning. Holding a Master's degree in Computer Science from Northeastern University, Hari is passionate about building state-of-the-art technological solutions. His expertise spans software development, web technologies, and cloud infrastructure, with a strong foundation in algorithms and full-stack development. Dedicated to driving organizational growth, Hari actively contributes to enhancing business outcomes through his technical skills and innovative approach to software engineering.
Scalable distributed systems, AWS Serverless API Gateway, machine learning, software development, web technologies, cloud infrastructure, algorithms, full-stack development, AI-focused software engineering
Currently working on AWS Serverless API Gateway, developing scalable distributed systems focusing on enhancing cloud infrastructure. Implementing machine learning integrations and web development solutions that optimize API performance and reliability. Leveraging technologies like Java, Python, and cloud-native architectures to build innovative serverless applications at Amazon Web Services.
Kevin Rohling RSVP Approved
AI Researcher/Engineer at Conifer Labs
# Community Analysis - Leverages customer conversations across social and CRM platforms, we implemented Discourse first for this hackathon
Sentiment Agent - This agent retrieves topic level posts and computes a quantified sentiment score for each topic as well as a rationale summarizing the topic conversations and why that led to the proposed score. Topic level scores are aggregated to compute a mean Community Sentiment Score
Idea Generation Agent - This agent leverages customer conversations to generate recommendations for product bug fixes and new feature ideas. New feature ideas are then passed to a PRD generation agent that generates documentation for new features that can be handed off to a PM for further refinement
Web Based Chat Agent (CopilotKit integration) - Embedded in our user interface is a chat agent that has access to all of the aggregated community data as well as agentially generated content (i.e. blog post, code examples, etc.). Users can query the agent about any data in our system and get immediate answers and recommendations.
System architecture/Stack:
NeonDB for Postgres database
PostgREST for API implementation
Vultr for API hosting
Python/Crew AI for agent implementation
TypeScript/Next.js for front end
Kevin Rohling is an AI engineer, researcher, and consultant focused on building practical AI systems that people actually use. He is a visiting researcher at UT Austin and earned his Master’s degree in Artificial Intelligence in December 2025. Kevin was previously Head of AI at Presence, where he led applied AI work across products and teams, and today he runs Conifer Labs, his independent consulting practice, partnering with companies to design and ship production AI systems. He is also a co-author of the paper “Factored Latent Action World Models,” which explores new approaches to scalable world modeling and embodied AI.
AI Agents, Reinforcement Learning, LLMs, Multi-Model/VLMs, Embodied AI
Agent Framework: https://bondai.dev/docs/intro
Jonny Tran RSVP Approved
Founder at Extralit
HeyDev implemented a multi-agent system using CrewAI’s Flow framework
The main flow orchestrated a series of specialized agents that work together to
Analyze Github Repository
Pulling CHANGELOG, PRs and Doc changes
Generate content topics
Create content draft
Review content for and publish to the database
Each agent in the flow has a specific role:
Repository Analyzer: Fetches and processes GitHub data
Content Strategist: Generates relevant topics
Content Creator: Writes engaging technical content
Content Editor: Handles post-processing and refinement
Jonny Tran is an open-source software engineer specializing in AI and LLM, currently working at health startup. With experience in research and software development, he has previously held positions at the Gates Foundation and Genentech. Nhat earned his Ph.D. in Computer Science, focusing on Graph Machine Learning & Bioinformatics, from The University of Texas at Arlington, where he also completed his Bachelor's degree. His expertise includes full-stack software development, machine learning, bioinformatics, and data engineering, and he actively contributes to open-source in scientific literature data extraction, and currently works in a clinical health AI startup as a founding engineer.
Neural networks
NLP
Graph machine learning
Semi-supervised learning
Recommendation systems
DNA/RNA sequencing
Sequence-to-function predictions
Bioinformatics
Open-Source Software
Extralit - a platform for LLM-based scientific paper data extraction that integrates human and model feedback loops for continuous LLM refinement and collaborative data extraction.