BUILD MODEGHLet's build
Backend Engineer
AI Systems
Bangalore, India

Bhargav Kacharla

Senior backend engineer with 10 years of experience, now focused on distributed AI systems, LLM agents, RAG pipelines, on-chain automation, and evaluation tooling. I build indexers, event-driven services, Kafka/Redis-backed workflows, WebSocket streams, and production systems where agents trigger real backend work.

Distributed Backend SystemsLLM AgentsIndexersEvent-Driven ArchitectureKafka / RedisWebSocketsRAG Pipelines
View featured work →GitHubContact↓ Download Resume
01 / SYSTEMS

Distributed backend systems for applied AI.

A fast scan of the event-driven, realtime, retrieval, evaluation, and agent execution surfaces I build.

[01]

Agent orchestration

Planning loops, routing, tool execution, retries, multi-step workflows, and observable agent traces.

[02]

Context engineering

Prompt pipelines, memory, structured project context, tool results, and compact task state for grounded agent work.

[03]

RAG pipelines

Ingestion, chunking, retrieval, reranking, answer grounding, citation flow, and hallucination reduction.

[04]

Vector databases

Embedding strategy, metadata filters, semantic search, similarity tuning, and retrieval quality debugging.

[05]

Structured outputs

Typed schemas, tool-call contracts, JSON outputs, validation, parsing, and deterministic handoff to backend services.

[06]

Evals

Retrieval and answer-quality evals, judge scoring, structured JSON validation, missing-context handling, and regression checks.

[07]

Distributed platforms

APIs, workers, queues, Postgres/MongoDB state, Kafka, Redis, Docker, cloud deployments, and production operations.

[08]

Indexers and streams

High-throughput indexers, event-driven pipelines, WebSocket updates, realtime state, low-latency monitoring, and replayable workflow traces.

02 / FEATURED WORK

Featured work.

Flash Agent
Independent AI Systems Project · GitHub-Native Coding Assistant
2025 — 2026

Built a GitHub-native autonomous coding assistant in Bun/TypeScript that orchestrates planner, implementer, and reviewer agents.

Turns a PRD into an implementation plan, code changes, test runs, and pull request reviews through a backend workflow rather than a single prompt.

Built Flash Discover, a repo-aware knowledge layer that ingests PRDs, plans, source code, and tests into Qdrant/Postgres.

Serves grounded, citation-based context to planner and reviewer agents through CLI and HTTP APIs.

BunTypeScriptMulti-AgentGitHubQdrantPostgresCLIHTTP APIsBackend
Knowledge Assistant
Independent AI Systems Project · RAG Assistant with Evaluation
2025 — 2026

Built a TypeScript/Bun knowledge assistant for product documentation with document ingestion, section-aware chunking, Qdrant vector search, and citation-grounded LLM answers.

Implemented incremental indexing with content and chunk hashes, stale vector deletion, and a Postgres document registry to skip unchanged docs and avoid unnecessary re-embedding.

Built retrieval and answer-quality evals with judge scoring for faithfulness, relevance, citation correctness, structured JSON validation, missing-context handling, and per-stage trace timings.

TypeScriptBunQdrantPostgresRAGIndexingJudge ScoringCitations
Agent Execution Workflows
Nunchi.trade · Event-Driven Financial Operations
2025 — Present

Built agent execution workflows for job discovery, task routing, transaction execution, retries, settlement, and failure handling.

Designed backend infrastructure for autonomous agents to trigger, validate, and complete product-critical workflows with low-latency monitoring and human-review paths.

Built trading and execution systems including matching engine, risk engine, liquidation flows, event-driven services, and real-time data pipelines.

Developed high-throughput indexing pipelines processing up to 1000 TPS for near real-time transaction and workflow visibility.

Agent WorkflowsEvent-DrivenIndexersTask RoutingRetriesSettlementWebSockets1000 TPS
Backend Indexing & Execution
Novastro · Distributed Data Systems
2024 — 2025

Built in-house indexers for EVM chains using TypeScript, GraphQL, and MongoDB to stream, process, and query on-chain data for analytics.

Implemented account abstraction and backend infrastructure for transaction execution pipelines.

Built internal services to coordinate distributed transaction execution across bundlers and relayers with job scheduling, retries, batching, Redis-backed state, Kafka event flows, and realtime updates.

Increased transaction processing throughput from 100 TPS to 1000 TPS.

TypeScriptGraphQLMongoDBRedisKafkaWebSocketsERC-43371000 TPS
03 / TIMELINE

Experience arc.

Independent Projects · AI Engineer

2025 — 2026

Built Flash Agent, Flash Discover, and RAG/evaluation systems using Bun, TypeScript, Qdrant, Postgres, CLI/HTTP APIs, multi-agent orchestration, structured outputs, and judge-based quality checks.

Nunchi.trade · Senior Engineer (AI Systems)

2025 — Present

Built agent execution workflows, automated financial operations, matching/risk/liquidation systems, event-driven services, real-time data pipelines, WebSocket visibility, and high-throughput indexing up to 1000 TPS.

Novastro · Backend Engineer

2024 — 2025

Built EVM indexers, GraphQL/MongoDB data systems, account abstraction infrastructure, Redis/Kafka distributed execution services, and throughput improvements from 100 TPS to 1000 TPS.

Oddz · Blockchain Engineer

2021 — 2023

Developed staking infrastructure and automation systems for options protocol workflows, including expiry handling, staking rewards, and dynamic risk adjustments with Chainlink Keepers and Gelato.

Rakuten · Data Engineer

2020 — 2021

Architected data lake infrastructure for telecom data science workflows and evaluated big-data systems for petabyte-scale storage with MinIO and YugaByte.

OpenText / TCS · Software Engineer

2015 — 2020

Integrated content management systems with third-party APIs and optimized SQL Server stored procedures, reducing report generation time by 40%.

CONTACT

Building an AI backend or distributed system?

Reach out for backend architecture, coding-agent workflows, APIs, workers, queues, indexers, Kafka/Redis systems, WebSocket streams, eval design, RAG pipelines, or production AI integration.

Email meGitHub ↗X ↗LinkedIn ↗