SRM University, AP (11 May 2026- 13 May 2026) -
We are thrilled to announce the successful completion of yet another impactful Faculty Development Program (FDP), at the prestigious SRM University, AP.
This intensive 3-day program focused on Advanced AI & LLM Systems, diving deep into the technologies shaping the future of innovation. It was wonderful to engage with the brilliant faculty members and researchers at SRM University, exploring the practical applications and research potential of Large Language Models and Agentic AI.
At Global Infoventures, we are committed to bridging the gap between industry advancements and academic excellence. A huge thank you to all participants for their incredible energy and looking forward to many more impactful learning initiatives ahead!
Together, we are empowering the next generation of educators and innovators.
DAY 1
Setting the Stage for Innovation!
Day 1 of our Faculty Development Program (FDP) at SRM University, AP was a deep dive into the core of next-gen intelligence. We kicked off the series with a focus on Foundations of Modern Vision AI and Scalable Model Training.
It was an incredible experience working with the faculty and researchers as we navigated the complexities of high-performance AI. Here’s a snapshot of what we covered :
* Evolution of Vision AI
We explored the journey from real-time detection architectures to advanced multimodal frameworks. The focus remained on building systems with enhanced contextual understanding and superior decision-making capabilities.
* High-Performance Training & Optimization
Theory met practice as participants -
- Performed custom model training on domain-specific datasets.
- Leveraged high-performance GPU infrastructure and accelerated workflows.
- Implemented optimization strategies to improve inference efficiency and minimize latency for production-scale deployment.
* Scaling via Distributed Systems
To tackle the bottlenecks of large-scale AI, we delved into distributed GPU environments. From overcoming memory limitations to implementing data-level and model-level parallelism, we pre-trained generative transformer architectures to understand the blueprint of scalable AI system design.
The energy and technical curiosity from the participants at SRM University, AP was very encouraging for our team of trainers!
Stay tuned as we update on how we continued to push the boundaries of AI & LLM systems over the next two days.
DAY 2
Continuing the momentum at SRM University, AP!of our Faculty Development Program (FDP) was all about speed, efficiency and the practical side of massive models. We shifted our focus to Accelerated Transformer Workflows and Efficient Adaptation of LLMs.
As the demand for Generative AI grows, the ability to adapt these models sustainably is becoming a critical skill for both researchers and industry leaders. Here’s what we explored -
* Accelerated Transformer Pipelines
We moved beyond basic training to GPU-accelerated pipelines. By focusing on high-throughput workloads, we demonstrated how to achieve faster experimentation, training and inference — essential for modern deep learning tasks where time-to-market and compute costs are key.
* Parameter-Efficient Fine-Tuning (PEFT)
One of the highlights was our deep dive into lightweight fine-tuning strategies. Participants gained hands-on exposure to -
- Low-bit optimization and quantization.
- Instruction alignment to sharpen model accuracy.
- Domain adaptation techniques that maintain high performance while drastically reducing memory and computational requirements.
* Scalable Foundation Models
We addressed the "real-world" challenge : How do you deploy these models on limited hardware? From quantized training strategies to efficient checkpoint handling, we explored how to customize large generative models for specialized enterprise and research applications without needing a supercomputing cluster at every turn.
It’s inspiring to see the faculty and researchers at SRM University, AP engaging with these complex workflows to build deployment-ready AI systems.
Day 3
at SRM University, AP’s Faculty Development Program (FDP) : Delving into Intelligent Knowledge Systems and Autonomous AI Workflows!
The final session marked a massive shift from standard conversational AI to the future of enterprise automation — moving from standalone models to autonomous compound AI systems.
Here are the key takeaways and hands-on milestones from Day 3:
* 1. Intelligent Retrieval-Driven AI Pipelines Implemented context-aware pipelines enabling LLMs to interact seamlessly with external knowledge sources for more reliable, grounded and domain-aware responses.
Gained practical exposure to semantic retrieval workflows, vector-based knowledge indexing and real-time information integration.
* 2. The Rise of Autonomous Compound AI Explored how AI is transitioning into systems capable of complex reasoning, orchestration, planning and multi-step task execution.
Studied modular AI workflow design featuring interconnected intelligent components, dynamic execution pipelines and adaptive decision-making.
* 3. Hands-On Agentic AI Architecture Built and deployed agentic AI architectures using both memory-enabled and stateless execution flows.
Integrated prebuilt and custom tools to push model capabilities far beyond simple text generation — turning them into engines for intelligent task automation.
* 4. Observability & Reliability Deep-dived into execution tracing, debugging and real-time monitoring techniques to evaluate complex AI interactions and ensure production-grade reliability.
Wrapped up the FDP with incredible energy as we bridged the gap between theoretical AI and robust, autonomous workflows ready for real-world deployment!
After an extremely rewarding expeerience over the 3 days FDP, we look forward to many such interactions with the incredible team at SRM University, AP.
