I completed 3 years in VC in March this year. Year 3 was the toughest and most challenging yet. Here's what I learnt this year:
(if you haven't read Year 1 and Year 2 reflections, please see pinned tweet)
One of my favourite places to eat in London is the Jackalope where you have a pub with an amazing vibe, in the middle of a Mews and probably the best Xi'an cuisine in the basement. Absolutely banger noodles.
They have a similar sign! Also if you have not been to the Jackalope I recommend you go. Also the first sentence of this tweet is only possible in London. Undoubtedly the greatest city in the world.
Some of my favorite businesses are like Xi'an Famous Foods where the first-gen perfects the quality but the second-gen is operationally effective and takes scale to the next level. They are just very special places.
We first met Edra through a demo for our internal ITSM desk. Incredible team, let's see if the product is as great? Were we blown away - and more importantly, our head of ITSM offered to buy immediately and also referred it to other teams at EQT. Our operating teams jumped on and said this is what they were looking for in a number of their PE companies.
Today, I am excited to announce that we at @eqtventures are leading a $10M investment into @edra_ai which is creating the missing link for Enterprise AI adoption, which they call 'ContextOps'. Simply put, it is able to codify how your organisation *actually* works and not just what's in your runbook. Obviously, that unlocks a very different level of productivity for agents and that is evident in their rapid commercial adoption both at EQT and beyond.
Thrilled to be a part of this journey. Read coverage in the @WSJ in the next tweet.
@eqt started as a customer. They tested @edra_ai on their own service desk, then decided to invest.
Now, theyโre bringing us across their portfolio.
WSJ has the story:
@PaytonBarnwell I am not sure that folks might want to discuss and give bank access to the same agent/meta agent that also makes restaurant reservations for you for example. So I think 2-3 specialised agents and 1 general purpose one.
Thinking about personal assistants, I think there will be:
1 General purpose assistant (like instinct/muse)
1 purpose built assistant for every regulated and/or sensitive use case (like wealth, health, romance etc.)
Just watched Ramayana. 10/10 donโt miss it in the theatre. Jatayu fight sequence is incredible and sad. Yash as Ravana is an immaculate performance, soundtrack is great, @kapoorkkunal has great PMF as Lord Indra and the VFX I really liked! Canโt wait for Part 2.
@PaytonBarnwell I am not sure that folks might want to discuss and give bank access to the same agent/meta agent that also makes restaurant reservations for you for example. So I think 2-3 specialised agents and 1 general purpose one.
@PaytonBarnwell I am not sure that folks might want to discuss and give bank access to the same agent/meta agent that also makes restaurant reservations for you for example. So I think 2-3 specialised agents and 1 general purpose one.
@mikeyferg@Muse Yes exactly. You arenโt going to post stories on LinkedIn. Facebook (used to be) and Instagram (now) are the general purpose social network, LI specialised. Similarly you will have 1 general agent and 2-3 specialised.
@rishabhmjain Regulation + data sensitivity. Not referring to 'health coach' use case as much as 'medical assistant' to interface with the medical system
Thanks for your patronising comment, anon. My thinking was less about specialisation of agents, rather more about what users might want. In general when building products I have found it is helpful to think first from a lens of what users want, rather than thinking of whatโs technically possible unless you are inventing something new.