AI DSA Trainer (Algorithm)
Adaptive DSA tutoring with SLM-backed feedback and modular lesson routing (R&D intern production path).
SLMsGenAIFastAPIVectorDBMLOps
Overview
Algorithm is an AI DSA trainer built during the STSARC R&D internship — adaptive feedback, session intelligence, and modular lesson graphs.
Engineering narrative (structured layer)
The same constraints → decisions → tradeoffs → failures → evolution → impact pattern used for NeuroTicker can be added for this system in lib/systems.ts (constraints, tradeoffs, failureStories, evolution, proofMedia). Until then, use Systems architecture explorer on the home page for problem, pipeline, and engineering decisions.
Technical depth
- SLM pipelines integrated for low-latency tutoring loops.
- Context-aware routing APIs choose the next best module based on learner state.
- Session tracking ties outcomes to content variants for continuous improvement.
- Fine-tuned models deployed with measurable gains in resolution quality.
What recruiters should probe
- How routing decisions are logged and A/B tested.
- Evaluation harness for SLM outputs vs. human baselines.
- Cost and latency budgets for inference at scale.