PinnedInLevel Up CodingbyHari Ohm Prasath·Jan 7Build Self-Learning Agents Without Any Fine-TuningBuild agents that get smarter over time using dynamic tools and evolving system prompts — no fine-tuning required.A response icon1A response icon1
PinnedInLevel Up CodingbyHari Ohm Prasath·Jan 21Building Human-Like Agents with Right MemoryDesigning short-term and long-term memory for better agent experiencesA response icon2A response icon2
PinnedInLevel Up CodingbyHari Ohm Prasath·Nov 11, 2025Beyond Brute Force: Using Constraint Programming to Solve NP-Hard ProblemsA Step-by-Step Guide to Efficiently Solve Complex Assignment and Optimization ChallengesA response icon2A response icon2
PinnedInLevel Up CodingbyHari Ohm Prasath·Sep 2, 2024Debunking 5 Common Misunderstandings About Spark Dynamic Resource Allocator (DRA)Unveiling the Realities of Resource Management in Apache Spark
Hari Ohm Prasath·Apr 22How I Avoid AI “Dark Code”A staff engineers guide to using multi-model reviews and rigorous planning to eliminate dark code
InLevel Up CodingbyHari Ohm Prasath·Apr 15Building Autonomous Long Running Agents — From ScratchEngineering a durable harness for planning, building, and independent QA verification
InTowards AIbyHari Ohm Prasath·Mar 31Why NO one uses AI Code review toolsWhy generic bots fail and how to build an AI reviewer that actually understands your team’s unwritten rules.A response icon2A response icon2
Hari Ohm Prasath·Mar 16Your LLM is Lying About Logs (And Burning Your Tokens). Here’s the FixTired of your AI confidently saying “No occurrence found”? Try this parallel chunking technique instead.
InTowards AIbyHari Ohm Prasath·Feb 28From Napkin Sketch to Roadmap: 10x Your Product WorkflowA five-phase, evidence-backed process to turn rough ideas into clear PRDs and Jira-ready epics — without endless review loops
Hari Ohm Prasath·Feb 19What It Means to Be a Software Engineer in 2026 (and Beyond)How I am adapting from writing code fast to thinking in systems, solving ambiguity, and leading with AI