ergara

Work

Selected systems and prior work

We walk through any of them in detail on a call, and where a system was built before Ergara, it says so.

Delivered and run by Ergara

Super Donuts · multi-outlet F&BLive pilot

GPS-verified store checklists for a six-outlet chain

Problem
Super Donuts runs six outlets. Head office had no reliable way to know whether a store had actually done its opening checks, or whether a photo showed what it claimed, where and when it claimed.
Previous process
Stores were checked through WhatsApp photos and follow-up messages, with nothing tying a photo to a particular store, time or place.
What was built
A daily checklist system. A manager can only start once GPS confirms they are in the store, and every photo is stamped with the store, the time and the location. Each completed section goes to head office for review.
Human role
The store manager completes the checklist on site, because checking a shop floor needs a person there. Head office reviews every section and approves it or sends it back.
Result
In pilot at one of the chain’s six outlets since May 2026, with the opening checklist live. We claim no outcome figures: nothing yet measures time saved or compliance.
Ownership and maintenance
Maintained by Ergara under an ongoing annual maintenance arrangement.

How the checklist runs

GPScheck instampedcapturehead officereviewor returnapprove

Also in progress

An inventory reconciliation and cash-handover system is built and independently reviewed, and has been validated against four days of the chain’s real sales data. It is in final validation ahead of rollout, and is not yet running in the stores.

Built at Koenig Solutions

Edtech and sales operations350K+ questions · 5,000+ prospects

AI, retrieval and outbound systems, built at Koenig Solutions

Prior professional experience. Vaibhav Aggarwal built these systems as an employee at Koenig Solutions, who own them. This is not Ergara client work.

Problem
Koenig’s self-assessment platform needed technical questions across 1,500+ courses, more than a content team could write by hand. Support had to dig through course documents to answer questions. Outbound and sales coordination were largely manual.
Previous process
A content team wrote questions by hand. Support searched course documents by hand. Outbound was one person writing and sending by hand. Leads, trainers and call slots were coordinated manually.
What was built
Four separate systems: a generation pipeline for technical questions; an internal retrieval tool over 8,000+ courses and their materials, for the support team; an outbound pipeline that searched, enriched and segmented prospects, wrote outreach for each account and classified replies; and an AI sales-coordination engine that assigned leads and trainers, proposed call slots across time zones, managed pricing and fed a live dashboard.
Human role
Support staff used the retrieval tool to find answers. Replies were classified, and the conversations that followed stayed with the sales team.
Result
350,000+ technical questions generated across 1,500+ courses, with consistency checks. 8,000+ courses and their materials indexed for support. 5,000+ prospects searched, enriched and segmented.
Ownership and maintenance
Koenig Solutions owns all four systems. They were built during Vaibhav Aggarwal’s employment there, and Ergara does not operate or maintain them.

Question generation

350K+ questions
1,500+coursesLLMgenerateconsistencycheck350K+questions

Course retrieval for support

8,000+ courses
8,000+coursesmaterialsindexsearchretrievesupportanswers

Outbound and reply classification

5,000+ prospects
prospectssearchsegmentenrichper accountwriterepliesclassify

Related service: B2B sales automation

Sales coordination

inboundleadtrainerassigntime zonesproposelivedashboard

Find out what you shouldn't be doing.

30 minutes, no pitch, and a written plan you keep either way.

Get your free AI Ops Audit