Hi, I'm Ashish
ML Engineer & AI Systems Architect. Currently pursuing an MS in Artificial Intelligence at Northeastern University (4.0 GPA). Specialized in Transformers, LLMs, Agentic AI, and production MLOps. Co-founder of Twinly, an AI productivity platform. Building intelligent systems that impact at scale.
About Me
I'm an AI/ML engineer and Master's student at Northeastern University (MS AI, 4.0 GPA) with extensive experience in building production-grade machine learning systems. My expertise spans Transformers, LLMs, Agentic AI, and distributed systems. Previously, I designed time-series fusion transformers and Graph Attention Networks as an ML intern at Eizen (Rakuten SixthSense), improving anomaly detection accuracy by 18% and reducing root cause analysis time by 60%.
As Co-founder & CTO of Twinly, I architected distributed microservices using FastAPI, Kubernetes, and AWS, orchestrating AI workers to automate task management across Gmail, Notion, and Slack. I'm passionate about building systems that bridge AI research and real-world impact—from edge-based multimodal assistants to production MLOps pipelines that scale to 100K+ daily evaluations.
Experience
Co-Founder & CTO
TwinlyMarch 2025 – July 2025
Boston, MA
- •Co-founded an AI-powered productivity platform automating task management across Gmail, Notion, and Slack using secure API integrations
- •Architected distributed microservices using FastAPI, PostgreSQL, AWS EC2/Lambda/S3, and asynchronous event-driven workflows, achieving 98% uptime
- •Configured OAuth2 and JWT-based authentication with CI/CD-enforced security checks, audit logging, and telemetry dashboards (Prometheus/Grafana)
- •Orchestrated and fine-tuned Phi-3 agentic AI workers with telemetry-driven feedback loops, improving task relevance and multi-step action reliability by 30%
- •Led cross-functional team of four engineers and designers through agile sprints, delivering MVP three weeks early
Machine Learning Engineer Intern
Eizen (Contractor for Rakuten SixthSense Cognitive AI)January 2024 – November 2024
Hyderabad, India
- •Designed and trained a time-series fusion transformer enhanced with Neural Architecture Search (NAS), improving anomaly detection accuracy by 18%
- •Developed a Graph Attention Network (GAT) to model cross-entity dependencies, reducing root cause analysis time by 60%
- •Assembled containerized inference services using FastAPI and Kubernetes on GCP, enabling 100K+ anomaly evaluations/day with robust CI/CD pipelines
- •Architected a multi-agent conversational system using LangGraph, orchestrating tool-using agents to retrieve and reason over real-time observability data
Admin & Digital Lead
TEDxVBIT
December 2023 – March 2024
Hyderabad, India
- •Led end-to-end event operations for TEDxVBIT, coordinating logistics, marketing, and speaker management for an audience of 100+ attendees
- •Designed and launched the official TEDxVBIT website from scratch using HTML, CSS, JavaScript, achieving a 25% increase in online registrations
- •Managed website content, implemented SEO best practices, and integrated event ticketing/payment systems
- •Collaborated with design and media teams to deliver branded digital assets, ensuring consistent visual identity across online and offline platforms
Programming Lead
coding.Studio( ); (Student Club)June 2023 – May 2024
Hyderabad, India
- •Led a team of 11 developers to design and build scalable web applications using TypeScript, React, and RESTful APIs
- •Guided architecture decisions and applied best practices in Git version control, code reviews, and CI/CD pipelines, improving code quality and deployment stability by 30%
- •Managed project planning, sprint execution, and stakeholder communication under Agile workflows, delivering features on time and meeting client requirements
- •Mentored junior developers in full-stack development and advanced debugging, strengthening team capability and collaboration
- •Directed codebase refactoring efforts, cutting technical debt by 25% and improving maintainability
Artificial Intelligence Intern
Digital Motives Technologies
February 2023 – June 2023
Hyderabad, India
- •Identified and curated high-quality training datasets through targeted online searches, ensuring data relevance and diversity
- •Implemented and trained 3 ML algorithms (Random Forest, Support Vector Machine, XGBoost), achieving F1 scores above 0.85
- •Enhanced existing ML frameworks and optimized model pipelines to improve accuracy and reduce training time
- •Performed model testing, validation, and tuning to ensure robust and reliable deployment-ready systems
SDE II | Full-Stack Development
coding.Studio( ); (Student Club)August 2022 – June 2023
Hyderabad, India
- •Designed, developed, and deployed scalable web applications using TypeScript, React, Node.js, and SQL, serving 7,500+ active users
- •Managed the full software development lifecycle — from feature planning and architecture to production deployment — achieving 95% on-time delivery across sprints
- •Implemented unit and integration testing strategies, improving pre-deployment bug detection by 30%
- •Optimized backend data processing logic and SQL queries, cutting data fetch latency by 40% in high-traffic modules
- •Collaborated with product managers and UI/UX designers in an Agile environment to deliver user-focused features with rapid feedback cycles
Generative AI Developer
EpsilonPi (Student Club)
August 2022 – June 2023
Hyderabad, India
- •Prototyped and benchmarked LLM-driven conversational AI systems using FLAN-T5 and HuggingFace Transformers, improving assistant usability by 20%
- •Systematized reusable pipelines for data preprocessing, fine-tuning, and deployment using Python and NLTK, reducing end-to-end prototyping time by 30%
- •Investigated dialogue generation behavior in LLMs, uncovering hallucination patterns and fine-tuning constraints
Penetration Testing Intern
TalaKunchi Networks Pvt Ltd
January 2022 – May 2022
Hyderabad, India
- •Conducted vulnerability scans on client systems, identifying and documenting 50+ security weaknesses for remediation
- •Performed penetration testing on networks and web applications to assess exploitable vulnerabilities
- •Analyzed scan results, prioritized security gaps based on severity, and worked with engineering teams to implement fixes
- •Utilized tools such as Nmap, Burp Suite, OWASP ZAP, and Metasploit to test security controls and strengthen system defenses
Featured Projects
Architected a modular agentic conversational system using ADK, enabling structured multi-agent coordination, contextual memory routing, and adaptive tool invocation. Integrated the Qwen large language model into multi-agent reasoning and tool-augmented workflows, improving dialogue accuracy by 25%.
Constructed a FaceNet-powered facial recognition module on Raspberry Pi, achieving 92% accuracy across 1,000+ real-world edge authentications. Integrated gesture recognition with OpenCV and multimodal fusion, improving lock reliability by 28% under varying lighting and occlusion conditions.
Built a voice and sign-language assistant achieving 87% command accuracy across 2,000+ test trials using TensorFlow and OpenCV. Automated 15+ system workflows for scheduling, application launching, and OS control, improving task completion efficiency by 25%.
Skills & Technologies
Specialized expertise in AI/ML systems, production infrastructure, and advanced techniques
Resume
Download PDF📄 Download my resume to view it, or open it in your browser's PDF viewer below.
💡 Tip: Click the download button above to save a copy of my resume to your device.
Get In Touch
I'm always interested in collaborating on AI/ML projects, discussing research, or exploring opportunities to build impactful systems. Feel free to reach out via email or connect with me on LinkedIn to discuss ideas!
Email (Northeastern)
gajjela.a@northeastern.edu
Email (Personal)
gajjela.ash@gmail.com
linkedin.com/in/ashish-gajjela
GitHub
github.com/ashish159565