Daily Digest
DONE
- Tasks done 24
- Drafts queued 12
- Awaiting review 3
- Flagged 1
All agents · healthy
I build AI agents & automations that do real work
AI Agent Engineer and co-founder at LFG Labs. I build agents that research, write, publish, monitor, check their own output, and report back — engineered to survive the real world, not just a five-minute demo.
5+
Years shipping software
3×
Companies co-founded
Node·TS
Core build stack
Experience across
● About
Building AI that does the work — reliably, on schedule, without a babysitter
I'm an AI agent engineer and co-founder at LFG Labs. I build AI employees: agents that run a process end to end — research, write, publish, monitor, check their own output, and report back — on a schedule, without someone standing over them. The goal is simple: take the repetitive work off your plate so your people spend their time on the parts that actually need a human.
I care more about reliability than demos. Anything looks great in a five-minute test; the hard part is surviving the real world at 3am — expiring tokens, rate limits, stale sessions, ballooning context, state drift. I design for those up front, with health monitoring, auto-restart, proper secret handling, and audit trails built in. Before AI, I spent 5+ years in Web3 security — auditing DeFi protocols and co-founding a security firm — and that adversarial, break-it-first instinct is exactly what makes my agents hold up.
● How I build
A build process designed to survive the real world
-
01
Map the process
Find the repeating, rule-based work that eats hours every week, and define exactly what "done" looks like.
-
02
Design for failure first
Expiring tokens, rate limits, stale sessions, context limits, refresh races, state drift — the failure modes that quietly kill automations, handled up front.
-
03
Build the agent
Node.js and TypeScript with LLM and Claude API integration, browser automation, and connections into the tools you already use.
-
04
Self-checking output
Quality checks that catch the usual AI filler, with a human approval step before anything goes live.
-
05
Operate & monitor
Health monitoring, auto-restart, secret handling, and audit trails — built to be run and maintained, not just handed over.
● Education & experience
The path so far
Present
Co-Founder & AI Agent Engineer · LFG Labs
Building AI employees — autonomous agents for sales, content, and operations that run end to end on a schedule. Sales agents that research, enrich, score, and draft outreach; content agents that draft, quality-check, and queue for approval; an orchestration layer that delegates tasks, verifies what comes back, and rolls it into one clear digest. Node.js, TypeScript, and the Claude API, engineered for reliability.
Oct 2025 — Present
Co-Founder · MegaFi
Co-founding and leading a MegaETH-based DeFi super-app: a custom DEX, a Concentrated Liquidity Manager with algorithmic rebalancing, and an on-chain options engine built for parallel execution and real-time settlement.
Jan 2024 — Present
Co-Founder & Security Lead · BlockGenesys
Lead security delivery for client engagements — risk assessments, the audit pipeline, and a team of security professionals held to a consistent, high standard.
Jun 2022 — Present
Smart Contract Security Auditor · ImmuneBytes
Full-cycle audits: recon, manual review, Foundry-based automated testing, Echidna/Foundry fuzzing, and detailed reporting.
Aug 2021 — Present
Smart Contract Security Auditor · BlockApex
DeFi protocol audits across the same pipeline — spec analysis, manual and automated review, and reproducible findings.
2020 — 2021
DApp Testing Engineer · XORD & VoirStudio
Test plans, cases, and bug reports for decentralized apps — automation with Selenium/Cypress, wallet and cross-browser compatibility.
2016 — 2020
BS, Software Engineering · University of Karachi
The engineering foundation underneath everything that came after.
● Selected work
Where I'm making impact
Building now
AI Agents @ LFG Labs
Autonomous agents for sales, content, and operations — they research, write, publish, monitor, check their own output, and report back in a single digest. Built on Node.js, TypeScript, and the Claude API, with an orchestration layer coordinating the fleet.
Sales & content automation
Agents that research, enrich, and draft outreach — and content that's drafted, quality-checked, and queued for approval.
Smart contract security
5+ years auditing DeFi protocols at ImmuneBytes & BlockApex — manual review, fuzzing, reporting.
● Services
How I can help
Sales Automation
Agents that research prospects, enrich them with company and contact context, score them against your ideal customer profile, and draft outreach that reads like a person wrote it. Multi-step follow-ups, sensible sending limits, and clean pipeline tracking — your team still closes.
Prospect researchEnrichment & scoringOutreach drafting
Content Automation
Agents that research a topic, draft content shaped for the platform it's going to, run it through quality checks that catch the usual AI filler, and queue it for your approval before anything goes live — built so the output stays varied instead of repeating itself.
ResearchPlatform-shaped draftsQuality checks
Operations & Orchestration
The quiet, scheduled work — health checks, data syncing, recurring reports, monitoring, routine audits — plus a coordinating layer that hands out tasks across agents, checks the quality of what comes back, and rolls everything into one clear digest.
Scheduled jobsMonitoringOrchestration
● Tech stack
The tools that drive sharper results
TypeScript / Node.js Primary build stack
95%
LLM & Claude API Agent reasoning & tools
92%
Agent orchestration Multi-agent coordination
90%
Browser & API automation Real-world integrations
88%
Solidity Smart contract security
80%
AI & agents
Claude APILLM integrationAI agent development
AI app developmentPrompt engineeringOrchestration
Languages & automation
TypeScriptNode.jsJavaScript
Browser automationREST & APIsWebhooks
Reliability & ops
Health monitoringAuto-restartSecret handling
Audit trailsSchedulingRate-limit handling
Web3 background
SolidityFoundryEchidna
SlitherSmart contract auditingBlockchain architecture
● Working together
Built to run when no one's watching
5+
Years shipping software
3×
Companies co-founded
24/7
Always-on automation
Node·TS
Core agent stack
"I care more about reliability than demos. Anything looks great in a five-minute test — the real test is surviving the real world at 3am. I build agents to keep running long after delivery, not to break the first time something upstream changes."
Kaif Ahmed
AI Agent Engineer · LFG Labs
● Contact