Ksk Koushik profile photo 1 on Tal

Ksk Koushik

Senior AI Engineer @Memorae AI || Software Engineer(AI agents)@Aunix || Ex AI Engineer @codemate.ai || Ex AI research Intern @Gradehive AI || || Ex Gen AI intern @HiDevs Community || Building AI Agents for impact✨

Hyderabad

1+ year of work experience

Memorae AI
Aug 20262 mosPresent

Remote

Building the a memory system that is as precise as your brain. Building Knowledge graphs that map users memories
Aunix AI
Jan 20268 mosAug 2026

• Architected an Agentic system called Insights Agent that can handle large scale databases with millions of rows. • Designed and Implemented research backed Rapid Prototypes and Proof of concepts , which became the core of Insights agent • A message bus state architecture maintains context across all agents , and supporting flexible agent to agent communications. • Reduced the latency of overall system by 50% by harnessing in-process execution and columnar storage using Duck db. • Insights Agent integrates with 15+ platforms like Zendesk , Salesforce , Github , Jira and data lakes like Big query etc.. and is capable of correlating data form multiple sources semantically together during root cause analysis. • Insights agent is made with a clear separation of Data Fetching + Insights gathering and Root cause analysis as other part. • Where it gathers all the require data store them in a Virtual Insights ledger where every insight is connected as a knowledge graph , and the Root case analysis engine uses it to think through explore and find the exact concise and proof backed answers for users query. • Unlike other data agents, Insights agent isn’t designed to provide metrics , it is natively designed to give the true cause behind every movement of any named or un-named KPI of a organization.
TheAgentic
Jul 20255 mosNov 2025

Bengaluru, Karnataka, India

TheAgentic
Jul 20254 mosOct 2025

• Built a AI powered CRM management tool called Journey360.ai , using a complex Multi Agent Orchestration system. • Journey360.ai integrates 10+ crm apps and 3 communication apps with intelligent context management across all apps. • A Neo4j powered knowledge graph built by me using langextract maintains context across all the agents seamlessly. • I followed a Hivemind network agent model designed to reduce context loss between agents and better communication between agents • Implemented a Deep Research agent powered by in house built headless browser that processes 100+ urls for each research • Played a key role in developing the, Developing the research agent where it leverages power of RAG pipeline built for search by me • Managed core feature modules of Agentic AI, including auth sessions for offline models, performance improvements, and robust parallel processing that instantly makes makes parallel tool calls with langraph ReAct agents.
CodeMate® AI
May 20254 mosAug 2025

• Developed the backend of a Multi Agent Full stack web development system with 80000+ daily active users • Added Agentic Integration of context management and knowledgebases using advanced Retrieval techniques and Re-ranking for the full-stack AI agent powered by LLM(Large language Models). • Designed and implemented the Retrieval Augmented Generation Pipeline for large codebase embedding using Qdrant. • Added robust handling of backend to work seamlessly even during large AI models Integration and heavy active user requests • Implemented subprocess routes to breakdown tasks, significantly improving load times and overall performance. • Played a key role in developing the, codemate build feature which offers a comprehensive suite of Agentic AI workflows using fastAPI backend and LangGraph to help users build their dream web apps. • Managed core feature modules of codemate Agentic AI, including auth sessions for offline models, performance improvements, UI library integration, and Prompting techniques.
Gradehive AI
Mar 20254 mosJun 2025

Greater Bengaluru Area

• Designed a complex AI Pipeline for a AI Agents powered github-app called Deepdocs and AI integrated developer tools. • Performed research on handling large codebases without loosing any essential information. • Integrated with current Retrieval Augment Generation(RAG) processes and and used crewAI for building agents that handle most of the processes in the DeepDocs. • Implemented multi agent orchestration systems using OpenAI , Claude models with langchains ReAct(Reasoning Agents) Agent. • Developed an automated workflow that automatically analyzes the source code repository after each change and updates the documentation website using AI to always keep it in sync using re-ranking and Knowledge graphs with RAG.

Ksk’s AI Stack

claude code
claude code
claude
claude
langchain
langchain
langgraph
langgraph
openai
openai
DSPy
DSPy

Education

Bachelor of Technology - BTech, Computer science and artificial intelligence

Vignan's Institute of Information Technology, Beside VSEZ, Duvvada, Gajuwaka,Vadlapudi (P.O)Pin-530049 (CC-L3)

20224 yrs2026

Join Ksk and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Ksk Koushik on Tal

Headline

Senior AI Engineer @Memorae AI || Software Engineer(AI agents)@Aunix || Ex AI Engineer @codemate.ai || Ex AI research Intern @Gradehive AI || || Ex Gen AI intern @HiDevs Community || Building AI Agents for impact✨

About

I am passionate AI engineer having lots of experience in building production grade AI applications and complex AI systems powered by Agentic AI workflows . I am currently working as a AI research Intern at Gradehive AI helping them build AI integrated developer tools that harness the power of AI agents and make devs life easy . I also worked as Ex Generative AI engineer Intern at Genzoic building custom AI soltuions that cut the costs in companies workflows. using RAG and businness optimization tools generated by me.

Location

Hyderabad

AI stack

claude code

claude

langchain

langgraph

openai

DSPy

Work experience

Senior AI Engineer at Memorae AI, Aug 2026 - Present · 2 mos, Remote, Building the a memory system that is as precise as your brain. Building Knowledge graphs that map users memories

Software Engineer(AI Agents) at Aunix AI, Jan 2026 - Aug 2026 · 8 mos, • Architected an Agentic system called Insights Agent that can handle large scale databases with millions of rows. • Designed and Implemented research backed Rapid Prototypes and Proof of concepts , which became the core of Insights agent • A message bus state architecture maintains context across all agents , and supporting flexible agent to agent communications. • Reduced the latency of overall system by 50% by harnessing in-process execution and columnar storage using Duck db. • Insights Agent integrates with 15+ platforms like Zendesk , Salesforce , Github , Jira and data lakes like Big query etc.. and is capable of correlating data form multiple sources semantically together during root cause analysis. • Insights agent is made with a clear separation of Data Fetching + Insights gathering and Root cause analysis as other part. • Where it gathers all the require data store them in a Virtual Insights ledger where every insight is connected as a knowledge graph , and the Root case analysis engine uses it to think through explore and find the exact concise and proof backed answers for users query. • Unlike other data agents, Insights agent isn’t designed to provide metrics , it is natively designed to give the true cause behind every movement of any named or un-named KPI of a organization.

AI Developer at TheAgentic, Jul 2025 - Nov 2025 · 5 mos, Bengaluru, Karnataka, India

Artificial Intelligence Engineer at TheAgentic, Jul 2025 - Oct 2025 · 4 mos, • Built a AI powered CRM management tool called Journey360.ai , using a complex Multi Agent Orchestration system. • Journey360.ai integrates 10+ crm apps and 3 communication apps with intelligent context management across all apps. • A Neo4j powered knowledge graph built by me using langextract maintains context across all the agents seamlessly. • I followed a Hivemind network agent model designed to reduce context loss between agents and better communication between agents • Implemented a Deep Research agent powered by in house built headless browser that processes 100+ urls for each research • Played a key role in developing the, Developing the research agent where it leverages power of RAG pipeline built for search by me • Managed core feature modules of Agentic AI, including auth sessions for offline models, performance improvements, and robust parallel processing that instantly makes makes parallel tool calls with langraph ReAct agents.

AI Engineer at CodeMate® AI, May 2025 - Aug 2025 · 4 mos, • Developed the backend of a Multi Agent Full stack web development system with 80000+ daily active users • Added Agentic Integration of context management and knowledgebases using advanced Retrieval techniques and Re-ranking for the full-stack AI agent powered by LLM(Large language Models). • Designed and implemented the Retrieval Augmented Generation Pipeline for large codebase embedding using Qdrant. • Added robust handling of backend to work seamlessly even during large AI models Integration and heavy active user requests • Implemented subprocess routes to breakdown tasks, significantly improving load times and overall performance. • Played a key role in developing the, codemate build feature which offers a comprehensive suite of Agentic AI workflows using fastAPI backend and LangGraph to help users build their dream web apps. • Managed core feature modules of codemate Agentic AI, including auth sessions for offline models, performance improvements, UI library integration, and Prompting techniques.

AI Research Intern at Gradehive AI, Mar 2025 - Jun 2025 · 4 mos, Greater Bengaluru Area, • Designed a complex AI Pipeline for a AI Agents powered github-app called Deepdocs and AI integrated developer tools. • Performed research on handling large codebases without loosing any essential information. • Integrated with current Retrieval Augment Generation(RAG) processes and and used crewAI for building agents that handle most of the processes in the DeepDocs. • Implemented multi agent orchestration systems using OpenAI , Claude models with langchains ReAct(Reasoning Agents) Agent. • Developed an automated workflow that automatically analyzes the source code repository after each change and updates the documentation website using AI to always keep it in sync using re-ranking and Knowledge graphs with RAG.

Education

Bachelor of Technology - BTech, Computer science and artificial intelligence, Vignan's Institute of Information Technology, Beside VSEZ, Duvvada, Gajuwaka,Vadlapudi (P.O)Pin-530049 (CC-L3)

Public links

github: https://github.com/kskkoushik

leetcode: https://leetcode.com/kskkoushik

Ksk Koushik - Senior AI Engineer in Hyderabad | Tal